UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Determining the production and export potential for medium quality wheat using a sectoral model for Canada Webber, Christopher Alan 1986

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1987_A6_7 W42_4.pdf [ 9.24MB ]
Metadata
JSON: 831-1.0097201.json
JSON-LD: 831-1.0097201-ld.json
RDF/XML (Pretty): 831-1.0097201-rdf.xml
RDF/JSON: 831-1.0097201-rdf.json
Turtle: 831-1.0097201-turtle.txt
N-Triples: 831-1.0097201-rdf-ntriples.txt
Original Record: 831-1.0097201-source.json
Full Text
831-1.0097201-fulltext.txt
Citation
831-1.0097201.ris

Full Text

DETERMINING THE PRODUCTION AND EXPORT POTENTIAL FOR MEDIUM QUALITY WHEAT USING A SECTORAL MODEL FOR CANADA by CHRISTOPHER ALAN WEBBER B.Sc. The University of Briti s h Columbia, 1969 M.Sc. The University of Toronto, 1971 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE! REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Agricultural Economics We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December, 1986 ® Christopher Alan Webber, 1986 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 r»r cenat\ ABSTRACT In January 1985, the Canadian Grain Commission licensed a medium quality wheat cultivar, HY320, for production within the new classification "Prairie Spring Wheat". Field t r i a l s conducted between 1974 and 1985 have shown that HT320 outperforms Neepawa, a traditional hard wheat variety, by 25 to 30 percent on average. The objective of this study i s to examine the impact of this new high yielding variety on production opportunities and incomes for prairie grain farmers and to estimate the share of grain exports that i t may command in the future. The study w i l l also provide an estimate of the supply curve for HY320. A linear programming model i s developed i n which the country i s divided into 29 crop regions (22 of which are located i n the prairie provinces) and 7 livestock regions for which aggregate act i v i t i e s and constraints are defined. The a c t i v i t i e s can be divided into three major groups: produc-tion, shipping, and marketing a c t i v i t i e s . The model i s sector-wide in the sense that i t describes domestic supply and use of major crop and livestock commodities in Canada. The problem i s to determine the level of agricul-tural production which maximizes net returns to the agricultural sector subject to constraints facing the sector. Medium quality wheat i s i n i t i a l l y introduced into the model by allowing i t to compete directly with hard wheat for cropland allocated to wheat production. There i s also a limited capacity in the model for the new variety to replace other feed grains and oilseed crops. This constraint was later relaxed in the study. Prices of traditional crops were set at i i their 1984-85 level based on Thunder Bay. As l i t t l e medium quality wheat has been sold by the Canadian Wheat Board, there i s considerable uncertainty concerning i t s ultimate price. Consequently, the analysis was performed at eight specific prices between $135/tonne (the lower range for American and Australian medium quality wheat traded on the world market) and $170/tonne (the f i n a l realized price for HY320 in 1984-85). These limits correspond to prices of 0.72 and 0.91 relative to the 1984 blended price of high quality wheats ($l86/tonne). Results show that although total wheat acreages increase marginally over the price range, class composition changes dramatically. The percentage of medium quality wheat increases from 10? at $130/tonne to 94? at $170/tonne. As the price rises, the new variety becomes profitable to farmers i n an increasing number of crop regions. The c r i t i c a l relative price for most regions i s between 0.78 and 0.83. When the price i s $l40/tonne, the new variety i s grown in a band from east central Saskatchewan to west central Manitoba. At $l60/tonne, i t i s grown in a l l prairie regions except in the brown s o i l zone of south-east Alberta and south-west Saskatchewan. As the price of the new wheat rises, total wheat exports increase, although the quantity of hard wheat exported declines. When the price i s $135/tonne, wheat exports, which are up by 5%f consist of 6% medium quality wheat and 94$ hard wheat. At $170/tonne, wheat exports have risen by 37%, and 98$ of these exports are medium quality. The results indicate that the adoption of the new wheat has a negative impact on the production and export levels of a l l other grains. Over the price range examined, the increase in net farm income of prairie grain producers varies from $9 million to $715 million. Clearly, i i i any estimate of income effects i s sensitive to assumptions regarding relative grain prices. Based on the price of U.S. and Australian medium quality wheat varieties, as well as on market share considerations, the author feels that the export price of Canadian medium quality wheat w i l l be at the lower end of the price range examined, possibly between 0.72 and 0.80 the price of hard wheat, implying gains between $9 million and $155 million for prairie grain producers. Finally, the results of the study have implications for wheat licensing arrangements, marketing strategies, the grain delivery system, the trans-portation rate structure, farm assistance programs, and the direction of future research. iv TABLE OF CONTENTS ABSTRACT i TABLE OF CONTENTS iv LIST OF TABLES v i i LIST OF FIGURES x ACKNOWLEDGEMENTS xi CHAPTER 1. INTRODUCTION 1 1.1 Background 2 1.2 Problem Statement 8 1.3 Objectives . . 11 1.4 Procedures 12 1.5 Thesis Guide 14 2. THEORETICAL CONSIDERATIONS AND CONCEPTUAL MODEL 17 2.1 Some Theoretical Considerations . . . . . . 17 2.2 The Mathematical Programming Approach . . . . 21 2.3 The General Structure of the Canadian Model 23 2.3.1 The Crop Production Block 31 2.3.2 The Livestock Production Block 37 2.3.2.1 Beef Sub-block. 37 2.3.2.2 Hog, Dairy, and Poultry Sub-block 42 2.3.3 The Inventory and Crop Trade Blocks 43 2.3.4 The Livestock Trade Block 16 2.3.5 The Domestic Demand Block 51 2.3.6 The Objective Function 54 2.3.7 The Equations of the Model 55 2.3.8 Summary *>3 V 3. THE EMPIRICAL MODEL 64 3.1 General Description of the Data 65 3.2 The File Structure 67 3.3 The Crop Production Sector 69 3.3.1 Cost and Yield Data 69 3.3.2 Crop Rotations 73 3.4 The Livestock Sector 76 3.4.1 Diet, Cost, and Yield Data for Beef Production Activities 76 3.4.2 Production Information for Hog, Dairy, and Poultry Sectors 79 3.5 Shipping Routes and Costs 80 3.6 The Domestic and Export Market Sectors 83 3.6.1 Domestic Requirements 7ersus Export Levels 83 3.6.2 Commodity Prices . 84 3.7 Right Hand Sides and Structural Bounds 86 4. TESTS AND RESULTS 88 4.1 Model Validation 88 4.2 Incorporating Medium Quality Wheat 93 4.3 Results • 95 4.3.1 Base Case 96 4.3.2 Impact of Medium Quality Wheat on Production 98 4.3.2.1 Production Effects at the Provincial Level 98 4.3.2.2 Production Effects at the Prairie Level . . . .101 4.3.2.3 Production Effects at the Regional Level . 109 4.3.3 Impact of Medium Quality Wheat on Export Levels . . . . 114 4.3.4 Impact of Medium Quality Wheat on Income 118 v i 4.3.4.1 Change i n Net Farm Income 118 4.3.4.2 Change i n Export Earnings 121 4.3.4.3 Change i n Agricultural Sector Earnings . . . .123 4.3.5 Relaxing the Constraint on Medium Quality Wheat . . . 125 4.4 Modifying Assumptions and Limitations .129 5. SUMMARY AND CONCLUSIONS . 133 5.1 Summary of the Study . 133 5.2 Policy Implications of the Introduction of Medium Quality Wheat 140 5.3 Extensions and Modifications of the Model 143 BIBLIOGRAPHY 147 APPENDIX A. . Costs of Production of Major Crops for Western. Regions . . . 150 B. Yields of Major Crops for Western Region 154 C. Area Seeded to Each Major Crop for Western "Regions (Base Case) 158 LIST OF TABLES v i i 1.1 Summary of Four Broad Categories of Wheat and their Characteristics 3 1.2 Relative Import Prices of U.S. and Canadian Hard Wheats at Shipment Position Noted 4 1.3 World Imports and Volume Shares by Wheat Category 5 1.4 World Wheat Import Shares by Market Area (1971 and 1981) 6 2.1 Summary of Major Factors Affecting the Supply of Wheat . . . . . . 18 2.2 Crops Included in Model, Land Class Required, and Level for Recording Yield 33 2.3 Beef Animal Categories Specified i n the Model 41 2.4 Livestock Categories Used i n the Trade Block 47 3-1 Summary of Major Data Requirements of the Model 65 3.2 Summary of Input Data Files by Category . 68 3.3 Portion of Crop Data for Alberta, Region 1 71 3.4 Relative Yield Advantage of Medium Quality Wheat over Hard Wheat 74 3.5 Portion of Cropland Planted to Various Crops for Region 1 of Alberta (1979-1984) 75 3.6 Portion of Beef Data for Alberta 77 3.7 Cull Rates, Replacement Ratios and Calving Rates for the Beef Herd by Province 79 3.8 Livestock Coefficients Required for Hog, Dairy, and Poultry Sectors of Model - Alberta 80 3.9 Shipping Rates for Grain from Various Prairie Points to Export at Vancouver or Thunder Bay 82 3.10 Grain Prices Selected for the Base Run -1984-85 Average Prices 85 3.11 The 1984-85 Price for No. 1 CWRS and the Relevant Price Range for Triple-M Wheat Varieties 86 4.1 A Procedure for Model Validation 91 v i i i 4.2 Parametric Variations i n the Price of Medium Quality Wheat (MQW) 95 4.3 Model Results for the Base Situation - Area Planted to Each Crop by Western Province 97 4.4 Change i n Area Planted to Each Crop at Various Price Levels of MQW - Alberta, Saskatchewan, Manitoba 99 4.5 Change i n Area Planted to Each Crop at Various Price Levels of MQW - Western Canada . 102 4.6 Total Area Planted to MQW at Various Price Levels by Province . . . 103 4.7 Total Production and Exports of MQW at Various Price Levels . . . . 106 4.8 Acreage Response and Export Supply El a s t i c i t i e s for MQW 106 4.9 Total Area Planted to MQW at Various Price Levels by Region . . . . 110 4.10 Summary of Wheat Exports by Class at Various Price Levels of MQW 115 4.11 Summary of Increases i n Exports from Prairie Provinces of Major Grains with Introduction of MQW . . . . . . . . . . . . . . . . 117 4.12 Changes i n Net Farm Income of Prairie Grain Producers from Introducing MQW . . . . . . . . . . . 119 4.13 Change i n Export Earnings of A l l Grains from Introducing MQW . . . 121 4.14 Change i n Net Earnings of the Agricultural Sector from Introducing MQW . . . . . . . . . . . . . . . . . 124 4.15 Comparison of Base Case Results under Two Substitution Options for MQW 127 4.16 Change i n Area Planted to Each Crop i n Western Canada From Base #2 at Various Price Levels of MQW -Constraint on MQW Relaxed 1 2 8 A.1 Costs of Production of Major Crops by Region i n B.C 150 A.2 Costs of Production of Major Crops by Region i n Alberta 151 A.3 Costs of Production of Major Crops by Region i n Saskatchewan . . .152 A. 4 Costs of Production of Major Crops by Region i n Manitoba 153 B. 1 Yields of Major Crops by Region i n B.C. . 15^ B.2 Yields of Major Crops by Region i n Alberta 155 B.3 Yields of Major Crops by Region i n Saskatchewan 156 ix B. 4 Yields of Major Crops by Region i n Manitoba 157 C. 1 Area of Major Crops by Region i n B.C. - Base Case 158 C.2 Area of Major Crops by Region i n Alberta - Base Case 159 C.3 Area of Major Crops by Region i n Saskatchewan - Base Case 160 C.4 Area of Major Crops by Region i n Manitoba - Base Case 161 LIST OF FIGURES x 2.0 Outline Map of the Prairie Provinces Showing Crop Region Boundaries 25 2.1 General Structure of Model 28 2.2 Structure of Provincial Crop Production Block 32 2.3 Selected Portion of Regional Crop Production Block 35 2.4 Selected Portion of Provincial Livestock Production Block 3 8 2.5 Calf and Yearling Ranching and Feeding Alternatives . . 40 2.6 Selected Portion of Crop Trade Block 44 2.7 Selected Portion of Livestock Trade Block . . 48 2.8 Selected Portion of Domestic Demand Block . . . . 52 3.1 Overview of Matrix Generator and Input Database for Model . . . . . . . . . . . . 70 4.1 Acreage Response Curves for Medium Quality Wheat 104 4.2 Supply Curves for Medium Quality Wheat - Western Canada 107 4.3 Prairie Regions Adopting MQW at $l40/tonne 112 4.4 Prairie Regions Adopting MQW at $150/tonne 112 4.5 Prairie Regions Adopting MQW at $l60/tonne 113 4.6 Generalized Soil Map of the Prairie Provinces 113 x i ACKNOWLEDGEMENTS The model described in this thesis was the result of a team effort. The principal members included John Graham, Kurt Klein, and Geoff Paul. I very much appreciated Kurt Klein's invaluable knowledge of prairie agriculture and Geoff Paul's care in maintaining the database. I would l i k e to specially thank my major supervisor, John Graham, for his personal commit-ment to the project, and for his enthusiastic encouragement and generosity. It was a pleasure to work with a l l three people. I am grateful to Tim .Hazledine and B i l l Schworm for their suggestions and moral support, as well as for their care in reviewing the fi n a l draft. The research also benefitted significantly from critiques submitted by a number of economists with the Marketing and Economics Branch of Agriculture Canada which provided the funding for the project. Finally, I give special thanks to Shirley Stuible for her ideas, willing assistance, and encouragement at a l l stages of the thesis, and to my parents who know the value of education. CHAPTER 1 Introduction Medium quality wheat varieties, such as HT320, represent a new opportunity for Canadian wheat producers who currently face world market prices that many view as being depressed. Wheat i s not a homogeneous product and market trends over the last decade show medium quality wheat and soft wheat sales growing relative to that of Durum and hard wheats (Henning, Loyns and Carter). Regulatory agencies i n Canada have been reluctant to license these new varieties because i n most cases they cannot visually be distinguished from higher quality wheats and thus a mixing of wheats could hurt Canada's reputation as a renowned supplier of a high quality product. Canada i s able to command a premium price for i t s high quality wheats and yet farmers contend that the higher yields of medium quality wheats more than offset this premium price. Farmers have therefore exerted pressure on regulatory agencies i n Canada to permit them to grow varieties of wheat that are higher yielding but fetch lower market prices. One medium quality wheat cultivar, HT320, i s visually distinguishable and has been licensed for production since January, 1985. The objective of thi3 work i s to examine the impact that this new variety w i l l have on production opportunities and incomes for prairie farmers and to attempt to estimate the share of grain exports that this new variety may command i n the future. 2 1.1 Background The prominence of wheat in the Canadian economy i s indisputable. Grain i s one of the three major contributors to the Canadian balance of payments, and of grain and grain products, wheat constitutes 75 percent by value. On the world wheat market, Canada averages a 21? share of exports, being second only to the United States (Veeman and Veeman). With JH% of i t s wheat production exported, Canada must be keenly interested i n the dynamics of the world market. Wheat traded internationally can be classified into four broad categories based on physical characteristics and end use - hard, medium-hard, soft, and Durum. Table 1.1 provides the distinguishing properties of these four wheats along with information regarding major traders and end uses. As indicated, high quality (hard) wheat i s ideally suited for North American - West European loaf bread; the medium quality wheat for products such as unleavened bread, noodles, and chapatis popular in the Middle East and Far East; soft wheat for cakes and biscuits; and durum for pastas. Although the physical properties largely determine end use, the four categories are interrelated due to substitution p o s s i b i l i t i e s . The type of wheat grown in a particular region depends largely on climate and s o i l characteristics. Canada's growing conditions favour the production of hard, high protein wheat, and government policy has encouraged the development and production of this type through a s t r i c t licensing and inspection system. The result i s that Canada has earned the reputation of being a reliable supplier of uniform, high quality wheat and, consequently, top milling grades command premium prices. Table 1.2 presents the import price of an American hard wheat (U.S. Dark Northern Spring) relative to the import price of the comparable Canadian product (Canadian Western Red Spring) 3 Table 1.1: Summary of Four Broad Categories of Wheat and their Characteristics Category % Kernel Gluten Major Major Major Protein Hardness Strength Exporters Importers Uses Hard 11.5-14.0 Hard Strong Canada U.S.S.R. , loaf breads, U.S.A. China, pan de sal, Brazil, pan de agua Japan, U.K., Cuba, Netherlands Medium 10.0-12.0 Medium Medium U.S.A. U. S.S.R., noodles; Arab Australia China, steamed & Brazil, French breads; Japan, chapatis Iraq, Iran, Egypt, Israel Soft 8.0-9.0 Soft Weak France U.S.S.R., French breads U.S.A. China, pastries, Brazil, confection-Japan aries Durum footnote very footnote Canada Algeria, pasta, a hard a U.S.A. Italy, couscous, Tunisia bulgar a Durum i s a very .hard specialty wheat used for making pastas. Only the endosperm, that part of the kernel containing the protein, i s used. Sources: Ulrich and Furtan, p.6 Henning, pp. 47-49, 81-82 Grain Matters, June 1984 4 Table 1.2: Relative Import Prices of U.S. and Canadian Hard Wheats at Shipment Position Noted U.K. Netherlands Japan Period (CIF Tilbury) (CIF Rotterdam) (CIF Japan) 1971-72 to 1975-76 0.96 0.96 0.96 1976-77 to 1980-81 0.91 0.94 0.94 1981-82 to 1984-85 0.96 0.94 0.93 a Price of U.S. Dark Northern Spring divided by price of Canada Western Red Spring Source: International Wheat Council (various issues), World Wheat Statistics, London at three market locations (U.K., Netherlands, and Japan) during the period 1971 to 1985. Over the past 30 years, the world wheat market has experienced considerable structural change involving adjustments in the pattern of trade. Henning examined world imports by class of wheat from 1962 to 1981, and his data are summarized in Table 1.3. Wheat trade has increased at an average rate of 3.8? per annum, with medium quality wheat showing the largest gain by volume, although soft wheat had the highest growth rate. In contrast, high quality wheat trade increased at a modest 1.9$ per annum, giving up ground to the softer varieties. Loyns and Carter observed a similar pattern after comparing U.S. and Canadian wheat exports by class from 1975/76 through 1981/82. Two developments on the demand side have contributed to this trend. First, technological changes in the milling and baking industries (e.g. the chorleywood process, microwave ovens) have allowed the substitution of lower 5 Table 1 . 3 : World Imports and Volume Shares by Wheat Category Wheat Category Period Hard Medium Soft Durum Total 1962-66 Volume (Mtne) : 13.0 21.8 7.7 1.8 44.3 Share (%) : 29-3 49.3 17.3 4.1 100.0 1977-81 Volume (Mtne) : 17.0 32.6 18.1 3.7 71.4 Share (%) : 23.7 45.7 25.4 5.2 100.0 1965-79 Compound Growth Rate a: 1.9 2.9 7.3 4.4 3.8 a Growth rates are calculated by taking the mid-point3 of the 1962-66 and 1977-81 periods as the end points and assuming a constant rate of growth over the 14 year interval. Source: Henning, p. 60. quality (and thus lower priced) wheat without diminishing the quality of loaf breads. Second, there has been a shift away from traditional markets in Western Europe to Middle East, Far East, and centrally planned countries. The new markets have a preference for products favouring the softer varieties of wheat. The import shares of three market areas between 1971 and 1981 are compared i n Table 1.4. The decline in western European imports i s partly due to the technological changes referred to, and partly due to government policies which have encouraged grain production (Veeman and Veeman). On the other hand, population growth, increases i n per capita income, and favourable terms of credit have contributed to the higher imports i n the new market areas. In the case of the Soviet Union, there has been an increased emphasis on the domestic livestock sector at the expense of wheat for human consumption (Henning). 6 Table 1.4: World Wheat Import Shares by Market Area (1971 and 1981) Import Shares Market Area 1971 1981 Developed Market Countries 81? 19? Developing Market Countries 2? 43? Centrally Planned Countries 16? 38? Source: Veeman and Veeman, p. 29. There has also been pressure for introducing medium quality wheat from the supply side. The potential for substantial improvements i n yields of hard wheat appears to be exhausted. Additional gains must come from increasing the area planted, or sacrificing quality (Henning). Responding to these pressures, plant breeders i n Canada have been developing medium-hard varieties over the past 20 years, and experiments with these new cultivars show, substantial increases i n yield over traditional varieties. Despite these advances, the Canadian Grain Commission (the agency responsible for regulating the Canadian grading system) has been cautious about licensing the new varieties for fear of jeopardizing Canada's reputation for a high quality product, and the premium that product commands (Olrich and Furtan). As previously mentioned, this reputation rests on two p i l l a r s - f i r s t , the licensing system and, second, inspection procedures. Although farmers can grow unlicensed varieties, the product can only be used for livestock feed. Until recently, the only classes of wheat licensed were red spring, amber durum, red winter, soft white spring, u t i l i t y , and mixed grain 7 (Wilson). 1 within each class, the number of varieties i s small. For example, in 1973, 80$ of the hard red spring wheat planted consisted of three varieties - Neepawa, Manitou, and Thatcher (Henning). Of these, Neepawa i s dominant. For a new variety to be licensed, i t must not only satisfy certain varietal standards specified i n the Canada Grain Act, but i t must also be visually distinguishable from cultivars of another class, since the o f f i c i a l inspection of grain relies partly on visual characteristics. To meet this requirement, certain desirable agronomic properties may have to be compromised. The alternative would involve the development of more sophisticated inspection, grading, and handling procedures to avoid the contamination of higher quality shipments. One of the medium quality wheat cultivars developed in the early 1970's, HY320, i s visually recognizable. The kernel i s red and slightly larger than those of Canada's hard spring wheats. It i s a semi-dwarf variety with s t i f f straw that reduces lodging. More importantly, the yields are considerably higher than those for traditional varieties. In f i e l d t r i a l s conducted between 1974 and 1985, HY320 outperformed Neepawa by 25 to 30 percent on average (Klein et a l . ) . Responding to pressure, the Canadian Wheat Board began subcontracting for small amounts of HY320 for test marketing. In 1984, 100,000 hectares were grown under contract. Finally, in January 1985, HY320 was licensed by the Canadian Grain Commission under the new classification "Prairie Spring Wheat". In 1985, the number of hectares planted was 175,000, indicating that farmers are willing to sacrifice higher product price for the increased 1 Approximately 85$ of wheat acreage i s seeded to hard red spring, 12$ to durum, 2$ to red winter, and less than one percent to soft white spring wheat (Veeman and Veeman, 1984). 8 yields (Agriculture Canada news release, 1986-03-14). As seed becomes readily available, the area planted could increase considerably. Despite the attractions of HY320, i t has several undesirable properties, including a longer maturation period; poor resistance to root rot, loose smut, and bunt diseases; and susceptibility to sprouting (Gilmour). Furthermore, recent information indicates some d i f f i c u l t y with HI320 related to milling and baking (Canada Grain Council, 1986, P. 51). As other cultivars are developed with more favourable characteristics, the pressure on policy makers to license these w i l l continue. 1.2 Problem Statement The introduction of a completely new type of wheat creates a great deal of uncertainty among those involved i n the production and marketing of Canadian grains. Questions arise relating to both demand and supply. On the demand side, the important issues involve estimating the own price, cross price, and expenditure e l a s t i c i t i e s of demand for the different classes of Canadian wheat, both domestically and internationally, as these w i l l determine the effect of medium quality wheat production on the net revenue of producers. On the supply side, i t i s imperative to be able to judge supply responses of Canadian producers to prices established i n domestic and world markets. The focus of the current research i s on the supply issues. Henning recently completed an extensive study concentrating on the demand aspects of the world wheat market. He investigated the economic relationships between the four classes of wheat traded internationally -hard, medium-hard, soft, and durum. Elasticity estimates were used to construct an econometric simulation model of the world wheat market with the .9 aim of estimating the impact of different Canadian export levels of medium quality wheat on prices and export earnings. Thirteen importing regions, and an aggregation of these were identified. Own price, cross price, and expenditure e l a s t i c i t i e s were calculated for each of four categories of wheat, and for the aggregate. These e l a s t i c i t i e s varied widely by wheat class and between regions, being heavily influenced by size of class expenditures - a small share tends to yield a large el a s t i c i t y and vice versa. Aggregate own price e l a s t i c i t i e s are smaller and more uniform than for regional estimates, however the magnitudes are large by traditional standards (Henning, p. 268). These findings are important i n the context of the present study because not only i s supply el a s t i c i t y important, but export demand ela s t i c i t y i s equally important in determining the benefits from the introduction of the new crop. Henning also calculated the supply and domestic demand e l a s t i c i t i e s for four of the exporting countries i n his model (Argentina, Australia, France and United States; Canada being excluded). Considerable variation was noted in the estimated demand e l a s t i c i t i e s reflecting the differing opportunities for substitution. For example, the estimates for Australia, which mainly exports medium quality wheat, were small relative to those for the United States which exports wheat i n a l l four categories. A major objective of Henning's work was to evaluate the price and revenue effects of varying the class mix of Canadian wheat exports. The impact of increasing the proportion of medium quality wheat was examined with the composition and the level of Canadian exports being set exogenously. The results indicated significant price impacts i n a l l four categories, although the response was greatest for hard and medium-hard wheat. As the proportion of medium quality wheat i n total exports 10 increased, the price of medium quality wheat f e l l and the price of hard wheat increased. He also estimated that export earnings could rise by $100 to $250 million per annum, findings i n line with those of Dlrich and Furtan, and Loyns and Carter. Henning acknowledged a number of problems with his approach. Although the supply of medium quality wheat from Canadian production i s set exogenously, i t i s important that i t be endogenous as farmers respond to prices as set i n world markets. Also, although Henning's simulations do estimate the equilibrium effect of quantity exported on price, he i s unable to compute demand e l a s t i c i t i e s from these results as the ceteris paribus condition i s not met. However, most studies indicate that the own price ela s t i c i t y for Canadian wheat sold on world markets i s elastic, and therefore net revenue should increase with production. Henning noted that another serious limitation of his study was the simplified supply model for each exporting nation. His approach assumed a competitive market with trading occurring at a single world price for each class, with no intervention being possible. In the case of medium quality wheat he did attempt to examine a change i n Canadian government policy by analyzing the impact of a change i n growing regulations on other major exporting nations. In this instance trade gains to Canada were more than offset by reductions in earnings to the other exporters. Thus, i t i s important to be able to examine supply responses i n the context of trade between competing nations and their resulting retaliatory actions. Dlrich and Furtan have pointed out that econometric studies tend to produce long run estimates, but they note that the Canadian Wheat Board and Canadian Grain Commission are often concerned with the short run marketing problems. They note that the demand for Canadian medium quality wheat could 11 currently be inelastic because of the size of American stocks and there i s danger i n any attempt to substantially increase market share at the expense of other nations, because of the threat of price wars. "Thus the threat of retaliation by i t s competitors may lead a wheat exporting nation to perceive that the ela s t i c i t y of demand for i t s wheat i s elastic only in the market share range that i s acceptable by the wheat exporting nation with the biggest treasury" (Ulrich and Furtan, p. 30). One may conclude that although supply and demand e l a s t i c i t i e s are important, market strategies both in the short and long run are equally important in determining negotiated prices and market shares (Carter and Schmitz; Grennes et a l . ) . Given these considerations, the emphasis i n this work i s on estimating supply responses for Canadian medium quality wheat assuming world market prices are given. Canadian net exports are determined as the residual after domestic requirements are satisfied. 1.3 Objectives The broad objectives of this study are the development of a regional model for the Canadian agricultural sector, and the application of this model to analyze the economic impact of introducing medium quality wheat as a production opportunity in the prairies. Several sub-objectives are involved. (1) To develop a regional model of Canadian agriculture, recognizing the interactions between crop and livestock sectors, as well as the prevailing transportation rate structure. (2 ) To assemble a documented data base for the model consisting of technical relationships between inputs and outputs, resource constraints, and cost coefficients. This w i l l include cost and yield data by producing area for the crop and livestock sectors. (3) To verify and validate the model. (4) To introduce medium quality wheat as a crop option, and to estimate the supply response by crop region for various possible prices of the new variety. Conclusions w i l l be drawn concerning producing areas where the new technology w i l l l i k e l y be adopted. (5) To aggregate regional crop production so that provincial and prairie production and export estimates can be computed. These include: the acreage response for medium quality wheat, the change i n optimal crop mix, the change i n export levels of each crop, and the change i n income. 1.4 Procedures To achieve the objectives outlined above, the following procedures are followed: (1) Based on economic theory, the variables determining the supply of medium quality wheat w i l l be specified, and from these relationships a linear programming model of the agricultural sector w i l l be conceptualized and implemented. The model w i l l include production, transportation, and marketing ac t i v i t i e s , for both the cropping and livestock sectors, and have as i t s objective the maximization of returns to the farming sector. Several software programs w i l l be developed to f a c i l i t a t e the assembly and manipulation of data bases and results of the model. (2) Since the data requirements of the model are extensive, data from a variety of sources including Statistics Canada, Agriculture Canada, and provincial ministries of agriculture are required. 13 The base year i s generally designated to be 1984. However, in certain instances i t w i l l be more appropriate to use values from the five year period from 1979 to 1984. For the prairies, where 22 crop producing regions are defined, the National Farm Survey data base w i l l be used extensively to define the input/output coefficients of the act i v i t i e s of the model, as well as the right hand side values. (3) To lend credibility to the results, some evidence of the model's validity i s provided. The procedure for model validation outlined by McCarl w i l l be followed, and several of the tests described i n that paper applied. Although "model validation" i s a d i f f i c u l t concept to define, the emphasis w i l l be placed on validity in terms of proposed use, as opposed to trying to mirror "perceived reality" exactly. (4) In the base model, a set of alternative crop combinations w i l l be defined for each crop region. Each such combination specifies the proportion of cropland seeded to the various grains, and when the model i s solved the optimal crop mix w i l l be selected for each region. In determining the possibilities for medium quality wheat there are three options for introducing the new crop to the model. First, the crop alternatives could be adjusted to include the new variety; second, medium quality wheat could be allowed to compete with traditional wheat varieties for the cropland available for wheat; and third, medium quality wheat, high quality wheat, and barley could be allowed to compete for the cropland allocated jointly to wheat and barley. For the purpose of the study, the last two options w i l l be of particular interest. Although crop 14 regions are defined for the entire country, medium quality wheat w i l l be restricted to the western regions of Canada. The amount of substitution between the two varieties of wheat and barley w i l l depend on the relative prices of the three crops. Although good estimates are available for the prices of the two traditional crops, there i s considerable uncertainty about the future price of medium quality wheat. As a result, a range of l i k e l y prices w i l l be tested, thus generating a series of solutions. (5) Each solution w i l l specify the amount of land allocated to each crop by region, and these are aggregated by western province, and for the prairies in t o t a l . Given the base run results, information on the amount of land allocated to each crop, the change i n area planted by crop from the base, the percentage change and the change i n total income earned by crop w i l l be computed for each medium quality wheat price and substitution option. From these results, the supply response curves for HT320 can be determined, the impact of introducing medium quality wheat on the other crops produced can be evaluated, and the.benefits of the new technology can be estimated. 1.5 Thesis Guide The present chapter has described the problem setting, objectives, and procedures of the study. In Chapter 2, the theoretical basis of the supply response model w i l l be presented, and i t s structural characteristics described. Chapter 3 begins with an overview of the data requirements. The specific data items have been organized into tables ( n f i l e s n ) , and the structure of these f i l e s w i l l be explained. Special attention w i l l be paid 15 to the cost, yield, and price data for medium quality wheat. Chapter 4 i s concerned with the tests and results. An overview of the validity testing w i l l be given, followed by a description of the medium qualiy wheat scenarios which were examined. The results w i l l then be presented and discussed and, f i n a l l y , the major simplifying assumptions of the model w i l l be identified. The study w i l l be summarized i n Chapter 5 and policy implications drawn, along with recommendations for further research. Special mention needs to be made concerning some of the background and funding of this project. The thesis was developed i n connection with a study contracted by the Marketing and Economics Branch of Agriculture Canada. The contract involved the construction of a regional model of the Canadian agricultural sector, and the use of that model to analyze f i r s t l y , the impact of introducing medium quality wheat on the prairies, and secondly, the effect of a change i n the transportation rate structure. Four reports were prepared as a result of that study, and three of these have direct relevance to this thesis. The report entitled "The Structure of CRAM: A Canadian Agricultural Model" (Webber et al.) describes the theoretical model, as well as i t s empirical content and validation. A revised description of the theoretical and empirical model i s contained i n Chapters 2 and 3 of this thesis, and a summary of the validation procedure i s provided i n Section 4.1. The Technical Appendix (Webber et al.) contains computer lis t i n g s of the complete database, the matrix generator and report writer programs, the model in MPSX format (IBM), the solution to the model, and some related information. The reader i s referred to that document for technical details concerning the model. Finally, the report entitled "Medium Quality Wheat Production on the Canadian Pr i a i r i e s : An Economic Assessment" (Klein et al.) describes an earlier version of the analysis whioh focuses on regional effects, study, and was completed following model by Agriculture Canada during made to the database. 16 The current presentation extends that a period of acceptance testing of the which time a number of revisions were 17 CHAPTER 2 Theoretical Considerations and Conceptual Model In this chapter the supply response of farmers i s discussed in a theor-e t i c a l framework, followed by a presentation of the conceptual model develop-ed to simulate this response. Factors influencing farmers' decisions are discussed within the context of various methodologies that have been used to estimate these responses both at the farm and at an aggregate level. 2.1 Some Theoretical Considerations It is important that policy makers be able to judge farmers' responses to changes. From a theoretical standpoint the issue is f a i r l y clear and yet economists continue to encounter d i f f i c u l t i e s in the accurate portrayal of this response in either the long or the short run. Traditionally, both math-ematical programming and econometric methods have been followed and each approach has i t s own particular merits (Merlove, Colman). The choice of method is often influenced by pragmatic considerations such as data a v a i l -a b i l i t y , personnel, and the time available for the study. This section w i l l present variables important in the discussion of supply response for medium quality wheat, while the ju s t i f i c a t i o n for selecting the mathematical programming method w i l l be given in Section 2.2. With the licensing of medium quality wheat by the Canadian Wheat Board, farmers must now add an additional crop to their l i s t of production opportu-nities and determine i t s p r o f i t a b i l i t y . Field tests in many different loca-tions have indicated a substantial yield advantage over the more traditional varieties but this advantage does vary by region depending on climatic 18 and s o i l conditions. Higher yields are offset by lower product prices for medium quality wheat and, therefore, producers are faced with a typical profit maximization problem in which risk and other factors may also merit consideration. In aggregate, i t is the producers' solutions to this problem which w i l l determine Canadian output and export levels. Factors influencing the acreage response for a particular class of wheat may be grouped into four general categories: economic, technological, i n s t i -tutional, and stochastic (Miketinac). The major items within these categor-ies are summarized in Table 2.1. Table 2.1: Summary of Major Factors Affecting the Supply of Wheat Category Factor Sign of Partial Derivative 1. economic i) own price i i ) price of alternative crops i i i ) price of required inputs iv) transportation costs (-) (-) (+ or -) 2. technological i) expected yields per hectare i i ) quantity of inputs required per hectare i i i ) farming practices (eg. rotational constraints) (+) (-) (-) 3. institutional i ) size of carryover stocks (-) i i ) domestic and foreign government policies (+ or -) 4. stochastic i ) va r i a b i l i t y In expected yield (-) 19 Of the economic variables, the effect of prices on the supply of wheat is straightforward, although the role of expected prices is less so. Own price is clearly very important. For a new product such as HY320, there is considerable uncertainty regarding the ultimate price level. Various views on this price level w i l l be presented in Section 3 . 6 .2 . The major alterna-tive crops to medium quality wheat are traditional wheat varieties, other feed grains such as barley, and to a lesser extent the oilseed grains, including canola and flax. The major variable input costs include payments for f e r t i l i z e r , chemicals, fuel, machinery repair, and wages, and these influence relative p r o f i t a b i l i t y of different crops. In addition, a farmer faces the cost of seed, insurance, building repairs, u t i l i t y payments, and interest charges. As crop related charges increase, the quantity planted w i l l decline, ceteris paribus.. Generally, an increase in transportation costs should affect the acreage response adversely. However, i t must be recognized that the alternative crops would face the same higher shipping rates, thus the more valuable crops would be favoured. Medium quality wheat may be used domestically as a feed grain and thus the proximity to livestock herds or poultry operations could affect farmers' planting decisions. Technological changes are continually impacting upon agricultural pro-duction patterns and trade. Aside from the introduction of new crops and new varieties, yield improvements are occurring. As expected yields increase, a crop becomes more profitable.. As mentioned in the introduction, yield data for HY320 has been obtained from f i e l d tests conducted over the past decade. Although the average yield does vary based on climatic and s o i l conditions, generally there is a 20 to 30 percent increase over traditional varieties. 20 Some of the input3 required were mentioned earlier with respect to the prices of these inputs. An increase in the quantity needed would also have a negative effect on production. Although the requirements for HY320 should be comparable to those of traditional varieties, an allowance must be made for extra fuel, repair and maintenance costs associated with higher yields. In addition, the seed must be pretreated to reduce i t s susceptibility to certain diseases. Institutional factors affect production decisions. Large carryover stocks could be perceived as a reduced demand for the crop, and thus have a dampening effect on new plantings. Domestic and foreign government policies include a wide range of items reflecting the complexity of the world wheat trade. These include quotas, subsidies, and export provisions such as credit and food aid programs. Stochastic factors include those items which contribute to the v a r i -a b i l i t y in expected yield, and thus to the riskiness of growing a particular crop- They include such items as weather, pests, weeds, and disease. Assum-ing risk averse behaviour on the part of farmers, as the risk associated with growing a particular crop increases, the area planted should decrease. A number of researchers have found that the agronomic risk factors associated with growing HY320 are considerably greater than with traditional varieties (Ulrich and Furtan). These include a longer maturation period, lack of resistance to root rot, loose smut and bunt disease, and susceptibility to sprouting. Other varieties of medium quality wheat have been developed with more favourable agronomic properties, but none are as readily distinguishable from traditional varieties as HY320. This section has identified the major factors influencing a farmer's 21 seeding d e c i s i o n , and i l l u s t r a t e s the complexity of the problem. In model-l i n g supply response, sufficient d e t a i l must be included to adequately reflect the d e c i s i o n process. 2.2 The Mathematical Programming Approach The two main techniques used to model supply responses in agriculture are regression analysis (econometrics) and mathematical programming (Colman, Shumway and Chang). With studies concerning the impact of new technology, there is usually l i t t l e or no h i s t o r i c a l data to assist the model building. Under these conditions, the mathematical programming methodology proves very useful. Within a linear programming framework, two techniques have h i s t o r i c a l l y been used to model the agriculture sector — f i r s t , the price endogeneous sectoral model approach, and second, the representative farm approach (McCarl). In sector models, the country is divided into regions for which aggregate a c t i v i t i e s and constraints are defined. A prominent example i s the CHAC model of Mexican agriculture (Duloy and Norton). In contrast, the representative farm approach consists of many farm models, each reflecting the structure of a homogeneous group of farms. The individual farms are then iteratively solved for an equilibrium. The Aggregative Programming Model of Australian Agriculture (APMAA), using this design, involves more than 500 representative farms located in 56 regions (Walker and Dillon) while that of Bri t i s h agriculture specified 624 farms in 8 regions (Thompson and Buchwell). Both approaches have been used to analyze problems relating to the adoption of new technology. Within.the representative farm framework, some examples are: 22 1. the use of the APMAA model to investigate the impact of four specific improvements in technology, two in the sheep sector and one each in the beef and wheat sectors (Wicks and C r e l l i n ) , 2. the use of a LP least-cost feed mix model for a single representative farm to estimate the expected economic benefits of investment in two alternative research programmes to improve yields of cassava in Columbia (Pachico, Janssen, and Lynam), 3. the use of two representative farm models to analyze the effect of change in labour and machinery on the future size of farms in England (Walford). Examples of the sector model approach include: 1. the use of CHAC to estimate the effect of government policy on mechani-zation (Duloy and Norton), 2. the use of the CARD system at Iowa State University to analyze the potential future needs of agriculture for water under different tech-nologies (Heady). Each method has advantages and drawbacks. Paraphrasing McCarl, repre-sentative farm models adequately represent microeconomic farming constraints, but they are cumbersome to use, and the linkage between the farm level and the aggregate is d i f f i c u l t to achieve. On the other hand, sector models are highly aggregate, adequately describing the overall markets while not captur-ing the f u l l factor-product substitution p o s s i b i l i t i e s existing in a repre-sentation farm model. Often sector models yield poor estimates of regional crop mixes unless constrained by f l e x i b i l i t y bounds. McCarl proposed various modifications to address this shortcoming of sector models and in particular, addressed the question of crop rotations. A solution, and perhaps the most straightforward procedure, i s to define a set of alternative crop mixes based on histo r i c a l data. Each crop mix would specify the proportion of seeded land allocated to each crop and these mixes may be validated by representa-tive farm models which detail constraints facing farmers as the production 23 period proceeds. In doing so, attention is paid to timeliness in production and to agronomic considerations which often result in rotations, and these are modelled in a reduced form approach through specifying crop mixes in the model. Further details of the agricultural sector model developed in the current study are presented in the remaining sections of this chapter. 2.3 The General Structure of the Canadian Model The purpose of this section i s to present an overview of the model developed to represent Canadian-agriculture. Note that this description of the theoretical model is a revised version of the one presented in Chapter 2 of the associated report entitled "The Structure of CRAM: A Canadian Regional Agricultural Model" (Webber et a l . ) . The model i s sector-wide in the sense that i t describes domestic supply and use of a l l major crop and livestock commodities in the economy and represents these on a regional basis. It is a 3ingle period model with 1984 chosen as the base period, and i t may be recursively solved to examine changes over a planning period of interest. Factor and product prices in this i n i t i a l version are given exogeneously rather than attempting to endogenize prices with corresponding representa-tions of producer and consumer behaviour. The problem is therefore to determine the level of agricultural production a c t i v i t i e s which maximizes net revenue subject to constraints facing the production sector. The " a c t i v i t i e s " can be considered variables whereas the "constraints" are equations inter-relating these variables. The model contains approximately 1180 a c t i v i t i e s , and 725 constraints. There are five geographical levels in the model: a nation level, an east and west Canadian level, "provincial levels", crop region levels and the 24 export or shipping points. "Nation" refers to a l l of Canada, which is sub-divided into the two "zones", western Canada and eastern Canada. There are seven provincial groups: Brit i s h Columbia, Alberta, Saskatchewan, Manitoba Ontario, Quebec, and the Maritimes (Eastern group). The model u t i l i z e s twenty-nine crop regions, twenty-two of which are in the prairie provinces including seven in Alberta, nine in Saskatchewan, and six in Manitoba (Figure 2.0). Each of the other seven provinces i s modelled as a single crop region. Currently, there are two ports defined in the model and they are Vancouver and Thunder Bay. Regarding production a c t i v i t i e s , two major sets are specified consisting of those dealing with crop or forage production, and those dealing with livestock. As noted above, crop production blocks or sets of crop producing a c t i v i t i e s are defined for 29 regions, whereas livestock blocks are defined at the provincial level for 7 different regions. Principal categories of crops include grain, forage crops, pasture and other minor crops. The four livestock classes considered are beef, dairy, hogs and poultry. The beef production a c t i v i t i e s are modelled in a f a i r l y detailed manner but single a c t i v i t i e s represent the other livestock categories. The submatrices of the model may be grouped into the following four sets: those dealing with production, those dealing with trade or transporta-tion, those dealing with domestic demand and those dealing with inventories of grains. The production submatrices deal with crop and livestock a c t i v i -t i e s . Because of the spatial disaggregation required, 29 separate sub-matrices specify crop production a c t i v i t i e s whereas seven submatrices detail livestock production. The trade blocks in the model may be conceptualized as those dealing 26 with both interprovincial and international trade i n crop and livestock commodities. Activities model the shipment of grains from production regions to domestic markets and ports. Similarly, transport a c t i v i t i e s describe the movements of l i v e and boned beef, and pork between provinces and export points. The domestic demand blocks model Canadian demand for both crop and l i v e -stock goods. For some items, consumption levels are specified for eastern and western Canada, and these demands are further disaggregated by province using a set of population distribution weights. For other commodities, demand is treated at a national level without regional disaggregation. The inventory blocks account, for opening and closing grain stocks held at points of shipment. The level of these- are currently specified exogen-eously in the model. Concerning the rows of the models, four major types may be identified. There are constraint rows, commodity balance rows, accounting rows and ratio rows. Constraint rows are specified in the model to deal mainly with land a v a i l a b i l i t y and to specify opening and closing numbers for the beef herd. Commodity balance rows deal with supply u t i l i z a t i o n for each of the major commodity items following the general specification that use cannot exceed supply. Accounting rows are used primarily to keep track of cash costs, and ratio rows have been specified to allow for alternative crop rotations and to specify certain replacement requirements in the beef breeding program. Further details of these constraint rows are noted below: Categories Number of Constraints 1. Cash cost rows 2. Regional crop production rows 3. Provincial crop balance rows 8 569 63 (continued) 2 7 4 . Provincial livestock production rows 126 5. Port crop balance rows 1 4 6. National crop balance rows 4 7. National livestock balance rows 5 3. National livestock trading pool balance rows 4 9 . International livestock trade rows ' __5 798 Figures 2 . 1 through 2 . 8 (excluding 2 . 5 ) i l l u s t r a t e the structure of the model. The symbols used in the tableaus are defined as follows: j : submatrix in the izn row, j t n column position of a larger ' matrix, Bj : set of structural bounds associated with the j t h column set, C i ±. ' column or row vector in the i t n row, j t h column position of a ' larger matrix, dealing with costs or prices, Cj : row vector of objective- function coefficients in the j t h column position, Dj_ j : diagonal matrix in the 1 t h row, j t n column position, R-L : set of right hand side values associated with the i t n row set, superscript : a l l values of matrix or vector are non-negative, superscript : a l l value of matrix or vector are non-positive, a : positive coefficient, a : negative coefficient, b : structural bound. Figure 2 . 1 i l l u s t r a t e s the general structure of the model. The objec-tive function row accounts for a l l revenues and cash cost3 associated with the model's a c t i v i t i e s . Crop exporting a c t i v i t i e s are represented by submatrices Cg and C^Q in Figure 2 . 1 . These ac t i v i t i e s account for Structural Bounds c o c 4 • #» o n jr • - * * rt S * I " i r-1 i » 4 rt O A jr > • o = ? » •< • ft rr • M 3 «? •8 — -o C 0 • M * rt rt • 3 . 1? •S ' M r» » • : < • 0 n sr 0 1 1 ^ 1 1 » o n C • « r* a 3 < • • s sr 1 0 i »— 9 n * •a* o •1 0 rf » 3 e < a a S jr 0 1 3 I l o V a a y e .a rt • 1 a n r» 4 - • g n rt »» O s M »» a Kt »»* r • *+ Crop S o 3 S S S a I *• ?• V w M '» r* + ft* r \ w L l v a a c o c k jr, o w a» n r • r \ 'a> Crop • S m 9 3 S » f- 1 Wi * »* .*+ •> L i v e s t o c k J» O 1 w> r + Crop W i m a •> < «*> •» • 9 T n S" w - • * > >• •at •> A -O* L l v a a c o c k 9 •al X f o r t 1 (Uaatam) - o 3 -m jr m Pore 2 (Eaatarn) .jr • » *> •0 « Uaatuarti n O •H m jr ** e dT O * o s • Eaatuard •w »— w •* j» **• j r »» » W r * o s c n < IT • m M .A o •* *-M n z + Crop c A M r» 8 H VI H e o • W w ** w L l v a a c o c k * «*• * Crop a* « • L l v a a t o c k « #• .? n K • Crop X • a r* »* O a m a— r i o L l v a a c o c k o n •i a *• o o o 1* o 1A o © 1* 1 * tm 1* t» 1* w I* SO o VI X V c? o n o :£ —i 2 9 crop sales through both western and eastern ports. In addition, revenue'from livestock or livestock product exports is accounted for b y C ^ , and the submatrices C- j^ to C-^j represent domestic demands at either eastern or western provincial levels, or at a national level. Hence, these groups of ac t i v i t i e s account for income earned by the agricultural sector while C^ g i s an accounting column or vector that subtracts a l l production costs ( C j ^ i through Cit&) and a l l transport costs (C^ q, ^ and ^ ) . These transport costs are defined for each of the shipping a c t i v i t i e s in the model which include crops shipped to Vancouver or Thunder Bay, as well as beef and hog movements. For explanatory purposes, crop and livestock production a c t i v i t i e s in Figure 2 . 1 take place in two western provinces and one eastern province. Although each province's crop production i s disaggregated by region, regional detail is suppressed in Figure 2 . 1 to simplify the presentation. The struc-ture i s best understood by examining the links between each submatrix in column order and row order. For example, crop production a c t i v i t i e s i l l u s -trated by submatrix A 2 1 are linked by column to A^ Q ^ , and by row to A 2 2» kn n and D_ The link between A, . and A,. , i s through production <-i" 2 , 1 2 <-»-!• 1 U , 1 or yields. Some cropping a c t i v i t i e s generate production which is detailed at the regional or provincial level, while other a c t i v i t i e s generate production which is accounted for at a national level. Wheat production, for example, is balanced at the provincial level (where production is a row in submatrix A 2 1) with use represented by A 2 2 » A 2 9 a n d D 2 1 2 * T n u s » wheat i s used for livestock feed (A 2 2)» exported or shipped to other provinces (A 2 >g), or consumed domestically, ( D 2 1 2 ) « At the national level, A.- A_ n - and A,n _ account for the supply 1 U , 1 1U , 3 1U , 0 of certain crops that f a l l into a national pool, and the domestic demand or 30 sales of these are specified through submatrix D . A s with a l l account-10,16 ing identities, use must be less than or equal to supply. The logic of the structural specification of the livestock a c t i v i t i e s can be explained with submatrices &2 2' A3 2 a n d A n 2* Submatrix Aj 2 represents the livestock consumption of grain or forages produced by the cropping sector. Rows and columns within submatrix k^,2 a c c o u n t ?or other inputs to the livestock sector. They specify a c t i v i t i e s involving livestock numbers and the transfer of animals from one category to another with time, and deal.with livestock yields accounted for at the provincial level. Sub-matrix A- accounts for yields which contribute to the- national pool 11,2 where supply (A,, „, A,, „ and A,., c) i s balanced with use (D ). At the 11,2 11,4 l l , b - . 11,17 provincial level supply i s shipped to other provinces or to export (A- ), J , 11 or i s retained for meeting provincial demand (D, . 0 ) . The structure of the submatrices for the other provinces illustrated in Figure 2.1 i s similar to that detailed for the f i r s t province. Regarding grain trade, shipments from province to port are specified. For example, a flow of grain from province 1 to a western port (Vancouver) is illustrated by submatrices A2 g and Ag Q, and submatrices A^  ^ Q , Ag 10 and Ag ^  serve a similar function for eastbound shipments. Export a c t i v i t i e s can be defined for each port. There are also livestock trading a c t i v i t i e s that move livestock or l i v e -stock products between any two provinces, between provinces and a national pool, and to export. These are illustrated by A^  A,. ^  and A^  ^  which draw from, or add to, provincial supplies, and by A^ ^ which draws from or deposits into the national pool. Exports may originate from this pool, or from the provinces directly. 31 Bounds or right hand sides specify upper levels on certain constraints and are also used to specify activity levels for exogeneously specified vari-ables. In the livestock sector, the magnitudes of the dairy, hog and poultry a c t i v i t i e s are exogeneous. Furthermore, in this particular study, inventor-ies of grains, certain import or export levels of livestock and livestock products, and domestic demand levels are a l l exogeneous. Finally, the open-ing and closing stocks of beef animal classes are specified by right hand side values. In Section 5.3, suggestions for extending the model w i l l be presented. A number of the proposals involve endogenizing some of the variables which are currently specified exogeneously. Figure 2.1 has been discussed so that the reader may understand the general structure of the- model before specific a c t i v i t i e s or rows of the model are reviewed.. The discussion that follows draws upon this overview and links each of these components or submatrices to a more detailed specification. 2.3.1 The Crop Production Block There is one crop production block for each provincial group. In Figure 2.1, the crop block for province 1 i s represented by submatrices C^  ^ , A2 2» a*1** *io l * T h e 3 e matrices are expanded upon in Figure 2.2 showing more structural details. In this example, province 1 has three crop regions. From Figure 2.2, i t can be seen that the ac t i v i t i e s are s p l i t into two groups: f i r s t , those dealing with regional crop production; and second, those concerned with the transfer of the yield to the provincial level where i t w i l l be used for livestock feed and domestic consumption. Crop production can also be shipped from the provincial supply to port. (See Figure 2.1 and crop trade block). 5 rr t» •o •i 0 < 3? •a e» 7 •a » Provl • r» I— o Objec 0 a 9 o 9 o 9 a 9 9 pi r» •> f o O • r-1 w *• M •» a r-r-H < • •1 a •o n t n »i rt «i n n n -o o n M 0 3 %l § o •o o •e •o •O rf f> » c n rt ••» O .9 •O o n rt •» • o 9 O 0 10 rr rt o 9 • *• >» n »» Region 1.1 g Rg o n o < S f t 0 1 JTt M •a s» KI n Region 1.2 oT, w > w w n r + w Region .1.3 s . •-• o •a i o «»«• d* o Melghc Ave. Region co Province 1 n g •A O n K»+ Region 1.1 co Province 1 TJ JO W m rH o n s < Z 5 M J \ n **• Ok Region 1.2 co Province 1 A o * 1 •J n •4 Region 1.3 Co Province 1 i >» e i > Ki O •» "« •a •a »» 33 In Figure 2.2 province 1 has three crop regions numbered 1.1, 1.2, and 1.3. The four submatrices in the f i r s t column of Figure 2.2 represent region 1.1's production. Submatrix ^ records the cash costs, while A^  ^  specifies the input requirements, constraints, and yield at the regional level. Crop yield is also modelled at the provincial level in submatrix A^ l f and at the national level in submatrix Ag ^- Generally, those crops that can be export-ed are specified at the regional level, those which are used only for feed, at the provincial level, and the minor crops, at the national level. These minor crops are then sold on the domestic market as per submatrix D^Q ^  of Figure 2.1. The crops included in the model are lis t e d in Table 2.2. Table 2.2: Crops Included in Model, Land Class Required, and Level for Recording Yield Crop Land Class Yield Recorded Wheat (high quality) Wheat (medium quality) Barley Flax Canola Soybeans Corn grain Corn silage Other (west) Other (East) Stored Forage Pasture crop land tame hay tame pasture & unimproved regional regional regional. regional regional national regional provincial national national provincial provincial 34 The submatrices for crop region 1.1 (excluding Ag )^ are expanded upon in Figure 2.3* The rows consist of: 1. Input Requirements including cash costs (recorded at the provincial level) and land constraints, 2. Ratios constraining use of fallow land, 3. Ratios constraining the quantity of each crop planted, 4'., Regional crop yield, 5. Provincial crop yi e l d . The columns consist of: 1. Uses of various types of land, 2. Set of ratios controlling the use of fallow land, 3. Set of ratios controlling the mix of crops planted. There are four types of land in the model: cropland, land in tame hay, land in tame pasture, and unimproved land. The amount of available land of each type in a region is specified: as a right hand side variable. Each crop i s constrained to one land type.. Table 2.2 l i s t s a l l crops included, and specifies the type of land required as well as the spatial level at which the yield is recorded. In the four western provinces, cropland can also be summerfallowed. Consequently, a distinction i s made between a crop planted on fallowed land and the same crop grown on stubble. The set of "planted: fallow" columns ensures that a minimum amount of land w i l l be fallowed. A further restraint ensures that at least as much land i s returned to fallow as is taken out. The set of crop ratio columns specifies the allowable mix of crops in a region. These ratios relate to actual crop rotations used in the past. The yield, in tonnes per hectare, for a particular crop is specified by the corresponding coefficient in the crop yield rows. These yields have been adjusted to allow for seed use, and waste. The yields of wheat, barley, n ti O 0 •o < tl e~ a u 9 r* n a U 3 Z V M k O C > J •1 » 0 l» ? 3* C • < a » n .<* •fc • • r II "» o i »> • •» n n 0 ? -c •C f 1- • O f 3 D. M 0 1 io fc9 »- n 0 f 9 o . « »-*>* » r? * • 6 t-0 rt • 1— 9 0 0 n 9 to l» C. r» •-• 0 "1 t- n fc t-f 0 C •9 «0 "O fc n n 0 0 t -«0 «o 0 - -C 9 9 D> "1 "O fc » »— >— t- fc • »- 9 9 0 rt o c a 0 B D. O-»1 9 0. W _ B n l£ O P 9 0 in 9 ft n f fc • t-t- 9 ft M 5? M 0 0 n. < C t-n 9 n n f a 0 3 t-n 0 IA n 44 fc fc 0 lO fc K B B .fc .» ft ft 0 0 5 3 : fc fc 10 O 3 M fc i« r> rt r »> p 0. n » -i« . I 44 » W fc» t-• 1 fc» >-t- 1— fc* e* t— fc • ItQ Wheat on Fallow HQ Wheat on Stubble c in m w> 90 O r> f i 90 K O ^ » • r -0 • I •-I f fc* f 1— fc» »-t— fc » HQ Wlioat on Fallow HQ Wlieat on Stubble fcl fc* •fc* C fc" 1— •r-•t-fc fc Barley on Fallow barley on Stubble • I • 1 »— 1— fc-•r* •fc" fc fc Ounola on Fallow Conola on .Stubble w l i - l 1— fc Sumwcrfallow ••1 -\ -\ f >->— » fc fc Tone Hay i Creen Feud Tame Pasture UnloproveJ Land K H X so m Q s" r E <» *| fc •-» | fc fc | * •* 1 TO TO tt fc « w >c n ^ -t- u t~ t~ fc fc a 0 0 >-;9 rt u 9 * H n O 0 •»» . * . & , - » » A • 0 « t- t l fc|fc | fc | fc | fc|fc|»|fc| fc|fc|fc>fc> fc|fc|fc|fc| n SO t| B O 10 M «A «J • »~ m ft S 0 r? w 9 n fc 0 rt n t- »-* 0 •f- fc o o e o o •»» r* "1 1— •0 KJ e e e o o o o o o o o c • j« t" £ t- 3 3 9 b o. a. a 0 9 9 9 o rt H O ^ 11 s fc •» 9 « B > O; fc fc »« » *> K 9 » a • «Q -•4 0 33 t» »— <o t~ n o t& »— 9 * B 3 O t-K> 9 » 0 — K 9 • »fl I— »- o r 3 i -t— 0 n "0 t* •a KJ " f c n 36 flax, canola, and corn grain are recorded at the regional level. These crops are transferred to the provincial level for domestic use, or for shipment to port for export. There are three options for effecting this transfer in the model: 1. a weighted average can be transferred from a l l producing regions in the province, to the provincial l e v e l . The weights are based on the relative sizes of the regions. (This option i s represented by the fourth column of submatrices in Figure 2.2), 2. an amount can be transferred directly from any one region to the provincial level. This option requires one activity per region (See columns 5, 6, and 7 of Figure 2.2), 3. when~shipping grain to port, a proportion can be held back for provin-c i a l use-When allowing, the direct shipment of grain from a region to port, options 1 and 3 ensure that at least part of a region's production w i l l be used domestically. However, both of these options are constrained (option 1 more than 3 ) , thus option 2 allows the f l e x i b i l i t y to ensure the provincial requirements for feed use and domestic consumption are met. A higher trans-portation cost associated with option 2 ensures i t i s only used when absolutely needed. Although a l l three options were tested, in the fi n a l version of the model a l l regional production is f i r s t transferred to the provincial level using option 2, and then shipments to port are made from this provincial supply- The region to province level transfers are implemented in the crop shipment block, enabling a unique transportation cost to be associated with each transfer a c t i v i t y . This aspect is important 3ince the Crow transport rate structure and alternatives to i t have been examined using this model (Klein et. a l . ) . The sum of shipping costs from region to provincial level 37 and from province to port, equals the cost of shipping directly from region to port. The primary provincial use of certain grains and forage crops is for livestock feed. Diets are expressed in terras of: stored forage, pasture, and barley for beef and dairy animals; barley for hogs; and wheat for poultry. To ensure feed requirements are met, the following substitutions are possible at the provincial level: stored forage can substitute for pasture; high quality wheat can substitute for medium quality wheat; and medium quality wheat can substitute for barley. In addition, corn silage (recorded under the provincial supply of stored forage) and corn grain are converted to the barley equivalent. In a l l cases, appropriate conversion factors are: applied.- Note that none of these substitutions are illustrated in Figure 2 . 3 . To fully understand how the interaction between the crop and livestock sectors i s modelled, livestock production must be studied. 2.3 .2 The Livestock Production Block There i s one livestock production block for each provincial group. In Figure 2 . 1 , the livestock block for province 1 i s represented by submatrices C l 2 ' A 2 2 ' A 3 2 ' a n d A l l 2 ' a n d t n e s e a r e expanded upon in Figure 2 . 4 . The four livestock groups modelled are beef, hogs, dairy, and poultry. Each of these groups w i l l be discussed in turn. 2.3 -2 .1 Beef Sub-block The beef sub-block contains the greatest detail reflecting the complexity and importance of this industry. It can take over two years r t o v i a c i A L tur rsooccriai Oaaatag Stack C U a l a f Stack OTUEa f MVttlCIAL LIVESTOCK rMMXTlOS al J «aV W . t a x f i l l Hand Swia • r o v l a c a f r o d u c t l a n Coata t r o v l a c l a l r oraao Crop H U n c a »aatu*a Uhaac' l a r l a v * * • * 4 • a * • • • « • * * * * < a I 0 i * 7 o !••* Sraa d l n a Hard Caaatoo Raplacaaaata Stack s t o c k a r a r a a d l o c Calvaa r a a d l o t L. t a a r l l n c a <_ |» Baa' Opaalna. Stock) Salaaca Eaplacaaaat S a t l o SraadlAa, Hard Uaanad Calvaa Y a a r l l n g a C u l l a a l_ 7 » < 0 7 0 7 o 7 b 7 a m«- it taat C l a t l a g Stock t r a a d l a a Hard l a p l a c a a a o t a Stockara f a a d l o t Calvaa f a a d l o c t . »aarllaa« • _ * ^ |- Saaf C l o a l a s Stock I P r o v i n c i a l Livaaceck T l a l d Saaf Slauohtacad USC BaaC rack 1 i • a a I 4 a < a 7 a 7 o N a t i o n a l Uvaatock T l . l d 0 •««< V a a l M i l k • r o l l a r a [99a • _ < a 7 e I 0 7 a < 9 S t r u c t u r a l Sounda • 4 « ? • » • \> Figure 2.4: Selected Portion of Provincial Livestock Production Block 39 before a calf is marketed or joins the breeding herd, and during this period i t passes through a number of stages. The cycle is described in Figure 2.5. As indicated, calves born about March may follow one of three routes. Typically, a rancher w i l l either s e l l calves in the f a l l or retain them as stockers. Feedlots which buy calves in the f a l l w i l l attempt to finish these animals within a six or seven month period depending on feed and fat cattle prices. The stockers which have been carried over the winter w i l l be sold as light feeders, retained and finished, or pastured as yearlings. Pasture yearlings then become replacements or feedlot long yearlings. Referring to Figure 2.4, the rows"consist of: 1. Input requirements: 1.1.) cash costs (recorded at the provincial level), 1.2) forage (from provincial crop supplies), 1.3) pasture (from provincial crop supplies), 1.4) barley (from provincial crop supplies), 2. Beef opening stock, 3. Beef balance, 4. Constraints on number of replacements: 4.1) replacement ratio, 4.2) number culled, 5. Beef closing stock, 6. Beef yield: 6.1) number of beef slaughtered, 6.2) provincial ABC beef yield, 6.3) national D beef yield.. The columns consist of: 1. opening stock a c t i v i t i e s , 2. within year a c t i v i t i e s , 3. closing 3tock a c t i v i t i e s , The opening and closing stock is divided into five categories. The number of animals in each of these is specified as a right hand side variable . CM I u i r om sonnet HONIHS (t) siocuts o» motor CALftl OVU UINTCR (t) f*tru«i UAAIIMS o* riioioi VtAuiMCi oyu $mvic» U»i> unutHun M rutioT IONS ; HONTNt M r . * f U IU» JUM July Aug. S«pt. Oct. Ho*. 0«c. J * o . Feb. «Uf . Apr i l H»f Junm July Aug. t « p l . Oct. Nov. fttc. Jin. f t t . ru r . ILTAII CAIVH FltOiOl l f . t d C l l v c l on F l n U M n q * ^ C A l V t l | l U l i o n (1 M t » > of o»lr.l ilAUCHU* fIAIN AS SIOCURS 0 V » UIMTtA ' f i l l A i 11 cm ncocAS Ml gntrqy A t t Ion r«cdlo« i t t r l l n q t ^ ^ tAUtHHA Figure 2.5: Calf and Yearling Ranching and Feeding Alternatives Ul of the appropriate constraint. In addition, there are four "within year" groups, two of which are used as balance row categories. Each category can be subdivided by diet. The beef classes are summarized in Table 2.3. Table 2.3: Beef Animal Categories Specified in the Model Number of Beef Category Diets Use Cows 3 0/S and C/S category (row & column) Replacements 1 0/S and C/S category (row 4 column) Stockers 1 0/S and C/S category (row 4 column) Feedlot calves 4 0/S and C/S category (row 4 column) Feedlot long yearlings 1 0/S and C/S category (row 4 column) Feedlot yearlings 4 Within year category (row 4 column) Culled replacements N/A Within year' category (row 4 column) Weaned calves N/A Within year category (row only). Yearlings N/A Within year category (row only) The: annual diet of an animal consists of a mixture of forage, pasture, and barley. When a stage spans the end of a calendar year, the diet is s p l i t between the opening stock (o/s) a c t i v i t y , and the corresponding closing stock (c/s) column. The s p l i t is based on the relative duration of the stage in each calendar year. The feed is drawn from the provincial supply. A certain proportion of the opening 3tock cows is culled, yielding D grade beef. The remainder of the breeding herd, along with the opening stock replacements, produce weaned calves based on an average provincial calving rate. A c a l f closes the year as either a stocker or a feedlot calf. A stocker starts the next year as a yearling, and then is assigned to one of three groups: feedlot yearlings, feedlot long yearlings, or breeding herd replacements. The number of yearlings retained as replacements is a fixed proportion of the opening stock of cows. Those replacements in excess of the 42 closing stock breeding herd requirements are culled, yielding "ABC" grade beef. The slaughter of feedlot calves, yearlings, and long yearlings also produces high quality beef, although the yield per animal varies with the category. The coefficients in the closing stock constraints are adjusted to allow for a death loss of between one and three percent. The slaughter of feedlot long yearlings i s also adjusted for death loss. 2.3.2.2 Hog, Dairy and Poultry Sub-block The provincial hog, dairy, and poultry sectors are represented in Figure 2.4 by the last four columns of the matrix. Hog and dairy are model-led by one activity each, while poultry requires two a c t i v i t i e s representing broilers and layers. The fixed bound on each of these a c t i v i t i e s specifies the number of animal units in the associated animal category. The column for provincial hog production has three coefficients: cash costs, barley requirements, and yield per animal unit. The cash costs and yield are recorded at the: provincial level, and the required barley is drawn from the provincial supplies. The dairy activity i s closely related to the beef sub-block. Each dairy cow requires cash input, and feed consisting of forage, pasture, and barley which i s met from provincial supplies. In turn, the cow produces milk which w i l l be sold on the national market.. A certain portion of the dairy herd is culled, yielding D beef. The remaining cows produce calves, based on a specified calving rate (not necessarily equal to the beef rate) which has been adjusted for the number of replacements required. A calf i s either slaughtered as veal, or is sent to the beef feedlot. The latter option is modelled by incrementing the "weaned calf" supply in the beef sub-block. 43 The provincial poultry sector consists of broiler and egg production. The poultry in each activity require cash input, and wheat drawn from the provincial supplies. Having discussed the production a c t i v i t i e s , the modelling of the ship-ping a c t i v i t i e s i s described next., Referring to Figure 2.1, the shipping or trade group consists of a crop shipping block and a livestock trade block. These two blocks w i l l be reviewed separately. Because the inventory block i s small and closely related to the crop trade block, i t w i l l be explained with the crop trade block. 2.3.3 The Inventory and Crop Trade Blocks There i s one inventory block for each of the two ports, Vancouver and Thunder Bay. They are used to specify the opening and closing inventories of certain, export grains, primarily wheat and barley. These are represented by submatrices Ag 7 and Ag g in Figure 2.1. The submatrix for Thunder Bay, Ag g, is expanded in Figure 2.6. In Figure 2.6, the inventory block occupies the f i r s t four columns where the f i r s t two a c t i v i t i e s represent the opening and closing levels of high quality wheat, and the second pair represents the opening and closing levels of barley. The rows of the inventory block consist of the Thunder Bay crop, balance rows. There i s one coefficient for each ac t i v i t y . The value of the coefficient i s "-l" for opening stock, representing a supply of grain, and " • f l " for a closing inventory, corresponding to a use. The level of each act i v i t y , in tonnes, is specified as a structural bound. As a l l crop shipments pass through one of the ports, the inventory blocks are closely associated with the crop trade block. The crop trade THUNDER. BAY INVENTORY SHIPPED FROM SASKATCHEWAN REGION 1 TO SASKATCHEWAN (PROV. LEVEL) SHIPPED FROM SASKATCHEWAN (PROV. LEVEL) TO THUNDER 9AY i SHIPPED FROH THUNDER BAY TO ONTARIO PROV. LEVEL) EXPORTED FROM THUNDER BAY TO WORLD Right Hand Side m Wheat at et s e e • a o o u Barley «o cs e c c M • o e. ~* o o • CJ x 5 .5 .» u m c* n ~i s a U. u a) •1 >. 5 Z M or «1 - i S O P . u <* a >» jl t> » - l . fat cr m s a IIQ Wheat Barley Flax Objective Function 1 • a • a C j 0 (rig. 1) National Transportation Coat* a a a a a a a a i 0 c i .to ( P l * - *> Sask. HQ Wheat Region I Bar l a / Crop Ylald Flax 1 1 1 < 0 7 0 I P » 4 > l 0 (Pig. 1) SasVc. Prov. HQ Wheat Level Crop aarley Balance Flax 1 r i 1 1 I <_ 0 • 1 0 7 b Ont. Prov. HQ Wheat Laval Crap Barley Balance Flax I I < o I ? *6,10 , r l « - 1 1 Thunder HQ Wheat Bay Crop Barley Balance Flax T i T i 1 I T 1 1 1 1 1 < o 7 o JLJ 1 0 (rig. i i 7 o Structural Bounds b b b b B a (rig. 1) j y B (Plg.l) Figure 2.6; Selected Portion of Crop Trade Block 45 block is depicted by the ten submatrices in columns 9 and 10 of Figure 2.1. Column 9 represents the movement of grain from two western provinces to Vancouver, whereas column 10 represents the shipment of grain from a western province to Thunder Bay, and from there to an eastern province or to export. The set of grains which are shipped includes high and medium quality wheat, barley, flax, and canola. The possible routes are from regions of surplus to areas of d e f i c i t . The prairie regions are surplus areas, while other provinces may have to import grain to meet domestic requirements. The routes include: 1. Alberta regions to Vancouver, 2. Saskatchewan regions to Vancouver, 3. Vancouver to B.C., 4. Vancouver to rest of world (export), 5. Saskatchewan regions to Thunder Bay, 6. Manitoba regions to Thunder Bay, 7. Thunder Bay to Ontario, Quebec, or Maritimes, 8. Thunder Bay to rest of world (export). Three provinces are used for demonstration purposes in Figure 2.1. Sask-atchewan i s the only prairie province which can ship grain west to Vancouver or east to Thunder Bay. For this reason, i t must be Province 2 i n Figure 2.1. Province 1 can be B.C. or Alberta, while Province 3 represents either Ontario, Quebec, or the Maritimes. Column 10 in Figure 2.1, representing the eastward movement of grain, is expanded in Figure 2.6. Specific place names are used in the detailed figure to c l a r i f y the example.. The columns consist; of four set3 of a c t i v i t i e s : 1. The shipment of high quality wheat, barley, and flax from region 1 of Saskatchewan to the provincial level, 2. The shipment of high quality wheat, barley, and flax from Saskatchewan (provincial level) to Thunder Bay, 46 3. The shipment of high quality wheat and barley from Thunder Bay to Ontario, 4. The export of high quality wheat, barley, and flax from Thunder Bay to . the world market. The rows consist of: 1. Transportation costs (recorded at the national level), 2. Saskatchewan region 1 crop production rows, 3. Saskatchewan provincial crop balance rows, 4. Ontario provincial crop balance rows, 5. Thunder Bay crop balance rows. • Normally, a transportation activity consists of three coefficients: a cash cost for shipping one tonne of grain along the particular route, a "+1" corresponding to the origin of the shipment, and a "-1" corresponding to the destination.. In the; case of export a c t i v i t i e s , there i s only one coefficient in the constraint portion of the tableau. This coefficient i s a "+1" corres-ponding to a use of the grain held at a port. However, each export activity does have an entry- in the objective function which specifies the price per tonne of grain exported. Thus, the revenue received i s recorded. The second subgroup of transportation a c t i v i t i e s i s included in the livestock trade block which w i l l be described next. 2.3.4 The Livestock Trade Block The livestock trade block is represented by the eight submatrices in column 11 of Figure. 2.1. Trade can take place between any two provinces, or between a province and the rest of the world. In the case of livestock, the "rest of the world" essentially refers to the United States. There is also a livestock pool which can act as an intermediate trading point. Any prov-ince, as well as the world, can trade with the pool. However, the quantity 47 shipped from the pool must be less than or equal to the quantity shipped into i t . The quantities shipped along specific routes are often bounded, reflect-ing known shipping patterns. However, not a l l routes can be identified, let alone bounds on the quantities involved. Therefore, the use of the livestock pool for certain commodities ensures each province's demand for livestock products i s met. The livestock trading block can be subdivided into two sub-blocks including one for cattle or beef, and the other for slaughter hogs or pork. There is trade in three categories of beef, and in two categories of hogs (Table 2.4). Table 2.4: Livestock Categories Used in the Trade Block Livestock Type Livestock Category beef feeder cattle (same as weaned calves) beef slaughter cattle beef dressed beef hogs slaughter hogs hogs dressed pork The structure of the beef and hog sub-blocks i s essentially the same. For demonstration purposes, a portion of the beef trade sub-block has been expanded in Figure 2.7. The columns are grouped into four sets: 1. Slaughter cattle trade a c t i v i t i e s , 2. Feeder cattle trade a c t i v i t i e s , 3. Dressed beef trade a c t i v i t i e s , 4. Net beef exports and imports. M c l 2 •o •» pi Q tl i - P.? P V? tc ft* s < n • fc* 6 fc < » • o> • £ C < « • X • n rt e » o> n fl • « r t o n « M r» i » a ° M • I I o o n S o « * *— o f t r -IT i r ir JV l! ., • i • m o o •— rl CA C <") I- A > 9 ! C b o a v K e Cj o A > at c won C# c C a V fc fc n . »> n • » n • fc A •C M n c 3 a , i D M A • r* ft A It ^ l l fk ft M n d 1 I I I A C CA I A A O •1 It fc 1 *-u • A > - 7 O. • It • n A I- 0. A 1- tfc O Oi • k c r> . » c r» • • « c n Ci — M l . <o fc «n f <a • f t A — a- i - » 7 H — sr i - 1— A £ < t I I < U n « » : H < o 'ft i t a <q » • •I M m • A A A W> A « p ia |1 1 W l l It M • A a . • A — C • A — (fc er * It- » -1 r> Sack, lu Out. n > C A :1 C i « c: n r j 9 » s'|fc« *• UorlU to Out. tr » -| « Saak. to UorlJ »i ~i C 52 o* * l •> t-| ft One. to World 14 •a* * Sack, to One. »-| - * Saak to Pool 71 ro t-| t- * One. to Pool u rn 9> •t— •-» * Pool to Sack. O t- Pool to One. 'A i - l • • I tr •a* * Sask. co World •tr •"1 * One. to World l m tr i-l •fc» •rn Pool to World - • Sask. to One. -1 - Sask. Co Pool o »-l •« * • One. to Pool ft c>> t A Pool to Sask. r n U •-• »-l Pool to One. ca m tr fc» -1 • - World to One. I*I t i XT »-| •- Sask. to World w r " G m XT fc. World to Puol tr I - l * Pool to World «• Feeders ro ••> Slaughter C. o f l «t| Ml ••! i ^ H 1— •> Dressed Beef IA t-1 • 1 Feeders •4 H • 1 Slaughter C. •5 R ie o ••• ro •-I ••1 Ilrcssed Beef t e H M l~ u> as » t-« f I U u X e e e o o o e o o e o o e a > *? n n • 1 *- »— i - t i — w »— M r— i — *— 1 -* - • * *— I-* — *•* -— —• ' • I •n <£ t - i * u i a i i — »• M I tu * • " i * 8^  49 The rows are also grouped into four categories: 1. Transportation costs (recorded at the national level). 2 . For each province, a subset of the beef balance rows representing: - weaned calves, - number of cattle slaughtered, - ABC beef (in tonnes). 3. For the national pool, a set of rows representing the balance in the following categories: - weaned calves, - number of cattle slaughtered, - ABC beef (in tonnes). 4. A set of rows to track the level of exports and imports in the following categories: - weaned calves, - slaughter cattle (in tonnes), - dressed beef (in tonnes). The feeder cattle and dressed beef trade a c t i v i t i e s w i l l be discussed f i r s t . A feeder cattle trade activity models the shipment of one feeder animal from the originating point (represented by "+1" in the weaned calf row of a province- or national pool) to the destination (represented by a "-1" in the weaned calf row of a province, the pool, or the world t a l l y ) . The cost of the shipment appears as a "•a" in the national transportation cost row. An activity in the third group of columns models the movement of one tonne of dressed beef from the point of origin ("+1" in the "ABC beef balance" row of a province, the pool, or the world tally) to the destination ("-1" in the appropriate "ABC beef balance" row). As before, the cost of shipping one tonne of dressed beef over the designated route is recorded in the transportation cost row. Slaughter cattle trade a c t i v i t i e s are slightly more complicated since slaughter animals are not modelled e x p l i c i t l y . Referring to Figure 2.4, which describes the beef production block, when a feedlot animal is slaughtered 50 the corresponding yield of beef is immediately added to the "ABC beef balance" row (represented by "-a"), and the total number of animals slaughter-ed is incremented by "+1". (The opposite signs are required since both con-straints are "less than" inequalities.) As a result, the shipment of one slaughter animal involves two rows for the point of origin, and two for the destination. Consider column 1 of Figure 2.7 which describes the shipment of one slaughter animal from Saskatchewan to Ontario. For Saskatchewan, the total number slaughtered is decreased by one, and the supply of ABC beef is reduced by the average dressed weight of a slaughter animal (the "+a" s i g n i -fies a use). For Ontario, the destination, the total number slaughtered- is incremented by "+1", and the: supply of ABC beef i3 increased by the equiva-lent amount of dressed beef shipped- The cost of the shipment, based on the average l i v e weight, appears as "+an in the transportation cost row-Note that in the set of accounting rows for the world, there is no count of the number of animals slaughtered. On the other hand, i t is necessary to distinguish between slaughter cattle and dressed beef where the units of both of these rows is "tonnes". This is because the selling price differs in the two cases. The fourth column section summarizes net external trade for the three categories of beef. If exports dominate imports then Canada receives income corresponding to the three positive coefficients in the objective function row. However, i f imports dominate, then a payment is required. Although export a c t i v i t i e s have been described for crops and livestock products, in the current specification commodities can only be sold on the world market once domestic needs have been met. The livestock feed require-ments were described in connection with crop and livestock production. A l l 51 other uses are included in the domestic demand block which w i l l be outlined next. 2.3.5 The Domestic Demand Block The domestic demand block is represented by the nineteen submatrices in columns 12 to 17 of Figure 2.1. As shown, domestic demand can be s p e c i f i -ed for western Canada, eastern Canada, or at the national level. Western demand (submatrices Dj 1 2» D3 13* D 4 , i 2 ' a n d D5,13^ d r a w s U P ° N t n e supply of the western provinces and, similarly, eastern demand draws upon the eastern supply. The shipping a c t i v i t i e s allow sufficient movement of goods to ensure domestic demand i s met.. As explained in the1 discussion of the crop and livestock blocks, the yields for certain minor commodities are recorded at the national level (sub-matrices A 1 Q 1 , A^ Q 2» Aio 5» An 2' A l l 4 a n d A l l 6^ * T o 3 I MP l i f*y t n e model, the entire supply of these commodities, with the exception of canola, i s sold on the national market (submatrices D^ g ^ and D^ JJ) yielding revenue (vectors C^ g and C ^ ) . A more detailed representation of the domestic demand block i s given in Figure 2.8. The demand for four commodities is specified for western and eastern Canada. They are high quality wheat and barley for the crop sector, and ABC beef and pork for the livestock sector. Upper limits on the demand for these goods are specified in the structural bounds row. This demand is distributed over the provinces based on population weights where the weight for a province i s the proportion of the population in that province. For example, consider the f i r s t activity in Figure 2.8 which models the sale of high quality wheat in western Canada. The domestic demand, specified by the © 1 s s » *r *< 9 « • tr e ft p-ft ft " t o : 3 5!? • * •—•ft t> ft 8 5 H M ft O Cf ft •W < ff *- n d 75 H rs-HQ I Barloy M C • M f fork S'S: " S E iBarlov ABC M l fork Caaola toykaam Othar Uaatarn Cra»« Otkar taacarn Crops B Boa' Vaal KB <zr-HUk BroUore o e c* c* e a> o e o |»>I» l"l» 10 53 "b" in the structural bound row, i s divided among the western provinces, two of which appear in the i l l u s t r a t i o n . The "a" in each "HQ wheat" balance row is the population weight for the associated province. The revenue resulting from the sale of one tonne of wheat is recorded in the objective function row. The residual, after domestic requirements are met, i s sold on the world market. The domestic uses of wheat include livestock feed and household consumption. The livestock in the model i s able to consume barley, medium quality wheat, or hard wheat. Because of these substitution p o s s i b i l i t i e s , the l e v e l - o r animal wheat consumption is responsive to price.- On the other hand, human wheat consumption is restricted to high quality varieties, and consequently, the price:-of medium quality wheat w i l l not affect consumers' decisions, concerning the level of wheat consumption. As the prices of a l l grains, other than medium quality wheat, are fixed in the current analysis, i t was decided to specify household demand exogeneously. In Section 5.3, several suggestions for extending the model w i l l be made including endogeniz— ing prices and domestics demand. The commodities sold on the national market include: canola, soybeans, "other" western crops, "other" eastern crops, D beef, veal, milk, broilers, and eggs. With the exception of canola, the levels of domestic demand are not specified for this group, and the trade in these commodities is not modelled. They are included in the domestic demand block s t r i c t l y for accounting purposes. Therefore their- unit prices appear in the objective function row under the column heading "national domestic use". The two "other" categories include a number of minor crops not e x p l i c i t l y modelled. In order to sum the production, their y i e l d i s recorded in terms of d o l l a r 54 value, and as a result, the objective function coefficient for each of these aggregate goods is simply "+1". Thi3 concludes the description of the a c t i v i t i e s and the constraints of the model. Because of the importance of the objective function in mathe-matical programming methodology, i t w i l l be discussed separately. 2.3.6 The Objective Function As illustrated in Figure 2.1, the model i s conceived as a large matrix such that the f i r s t row represents an objective function, and the remaining rows specify certain constraints.. Solving the model involves maximizing the objective function subject to the set of constraints. The objective function i s value added, or total revenue minus total variable costs. Revenue results from the sale of commodities on the world market (vectors Cg, C 1 0 and in Figure 2.1) or on the domestic market (vectors through C 1 < T). The coefficients in these vectors are unit prices and for selling a c t i v i t i e s are positive signifying revenue received. Costs arise from three sources: production costs, transportation costs, and the importation of certain commodities. The production costs (represent-ed by submatrices ^ through g in Figure 2.1) are recorded at the provin-c i a l level, while the transportation costs associated with the shipping a c t i v i t i e s are recorded at the national level (vectors C^g, C 1 > 1 0 and 3 ^ ) . These costs are summed (submatrix D-^  g), and subtracted from the total revenue (vector C^g). In the model, certain livestock items may be exported or imported (see Section 2.3.4 regarding the livestock trade block and Figure 2.7). Exports earn revenue, while imports increase costs. Consequently, vector C^ in Figure 55 2.1 is unsigned. As noted by Duloy and Norton, Hazell, and McCarl, alternative specifica-tions of the objective function may be appropriate under different circum-stances. The current specification was chosen given the problem structure and the fact that Canada may be viewed as a price taker in the markets that are of importance to this study. In the case of medium quality wheat, the model w i l l be solved for various prices of the new variety over a range considered to be reasonable. 2.3.7 The Equations of the Model The preceding sections have described the structure of the model with reference- to representative matrix tableaus in which the rows represent inequality constraints. Excluding the objective function there are 798 rows and, as explained in Section 2.3, they can be classified into the following nine categories: Number of Constraints 1. cash cost rows 8 2.. regional crop production rows 569 » 3. provincial crop balance rows 63 * 4. provincial livestock production rows 126 5. port crop balance rows 14 6. national crop balance rows 4 7. national livestock balance row3 5 8. national livestock trading pool balance rows' 4 9. international livestock trade rows 5 798 * _ not a l l . used As noted, not a l l rows are used in every generation of the model. The number depends on the options specified, but is generally in the order of 725. A statement of the equations of the system follows with the objective function being described last (see item 10). Number of con s t r a i n t s (1) Cash costs ( l . a ) National t r a n s p o r t a t i o n cost3 (TOTAL TRANSPORTATION COSTS) = (SINGLE REGION CROP TRANSFER COSTS) + (CROP SHIPMENT COSTS) + (BEEF SHIPMENT COSTS) + (HOG SHIPMENT COSTS) (l . b ) P r o v i n c i a l production costs (PROVINCIAL PRODUCTION COSTS) = (-PROVINCIAL CROP PRODUCTION COSTS) +• (PROVINCIAL LIVESTOCK PRODUCTION COSTS) ( 2 ) Regional crop production ( 2 .a) Land c o n s t r a i n t s 116 (AREA OF LAND CLASS i < (TOTAL AMOUNT OF LAND USED IN REGION) " CLASS 1 IN REGION) (2.b) Planted : fallow r a t i o s ( 2.b.i) Proportion planted 23 (PROPORTION OF CROPLAND PLANTED) < (SPECIFIED PLANTED ~ ' PORTION) ( 2 . b . i i ) Proportion fallowed 23 (PROPORTION OF CROPLAND FALLOWED) > ( 1 - [SPECIFIED PLANTED PORTION]) ( 2 . h . i i i ) Fallow balance 23 (PLANTED AREA PREVIOUSLY FALLOWED) < (AREA CURRENTLY FALLOWED) 57 (2.c) Crop mix r a t i o s 239 (PROPORTION OF CROPLAND PLANTED IN CROP o) (PROPORTION SPECIFIED IN SELECTED ROTATION) 145 (2.d) Regional crop balanoe (QUANTITY OF CROP c > PRODUCED IN REGION) (QUANTITY OF CROP TRANSFERRED FROM REGIONAL TO PROVINCIAL LEVEL) (3) P r o v i n c i a l crop balance 63 (QUANTITY OF CROP c > TRANSFERRED FROM REGIONAL TO PROVINCIAL LEVEL) + (QUANTITY REQUIRED TO MEET PROVINCIAL PORTION OF DOMESTIC DEMAND) + (QUANTITY TRANSFERRED TO NATIONAL LEVEL) + (QUANTITY SHIPPED TO PORT) (QUANTITY FED TO PROVINCIAL LIVESTOCK) (4) P r o v i n c i a l l i v e s t o c k production (4.a) Beef opening stock (NUMBER OF COWS FED AT START OF YEAR) (NUMBER OF REPLACEMENTS FED AT START OF YEAR) (NUMBER OF STOCKERS FED AT START OF YEAR) (NUMBER OF FEEDLOT CALVES FED AT START OF YEAR) (NUMBER OF FEEDLOT LONG YEARLINGS FED AT START OF YEAR) < (OPENING STOCK OF COWS) < (OPENING.STOCK OF REPLACEMENTS) < (OPENING STOCK OF STOCKERS) (OPENING STOCK OF FEEDLOT CALVES) (OPENING STOCK OF FEEDLOT LONG YEARLINGS) (U.b) Beef balance (U.b.i) Breeding herd balance ((CO'J/S RETAINED AT YEAR END) + (CULLED REPLACEMENTS) ] < (COWS AT START OF YEAR - CULLED COWS) + (REPLACEMENTS AT' START OF YEAR) ( 4 . b . i i ) Weaned c a l f balance I(STOCKERS PLUS FEEDLOT CALVES RETAINED AT YEAR END) + (FEEDER CALVES SHIPPED OUT OF PROVINCE)] < [(BIRTH RATE) * (COWS PLUS REPLACEMENTS AT START OF YEAR)] -t- (DAIRY CALVES TRANSFERRED TO FEEDLOT) + (FEEDER CALVES SHIPPED INTO PROVINCE) ( U . b . l l i ) Yearling- balance [(FEEDLOT YEARLINGS) + (REPLACEMENTS PLUS FEEDLOT LONG YEARLINGS AT END OF YEAR)1 < (STOCKERS. AT START OF YEAR) (4.c) Replacement r a t i o (REPLACEMENTS RETAINED AT YEAR END) > [(REPLACEMENT RATIO) * (COWS AT START OF YEAR)] (4.d) Constraint on number of replacements c u l l e d (REPLACEMENTS CULLED) < (REPLACEMENTS AT START OF YEAR) (4.e) Beef c l o s i n g stock (NUMBER OF COWS AT YEAR END) - (DEATH LOSS) > (CLOSING STOCK OF COWS) (NUMBER OF REPLACEMENTS AT YEAR END) - (DEATH LOSS) > (CLOSING STOCK OF REPLACEMENTS) 59 (NUMBER OF STOCKERS AT YEAR END) - (DEATH LOSS) > (CLOSING STOCK OF STOCKERS) (NUMBER OF FEEDLOT CALVES AT YEAR END) - (DEATH LOSS) > (CLOSING STOCK OF FEEDLOT CALVES) (NUMBER OF FEEDLOT LONG YEARLINGS AT YEAR END) - (DEATH LOSS) _> (CLOSING STOCK OF FEEDLOT LONG YEARLINGS) (4.f) Number of beef slaughtered (NUMBER OF BEEF SLAUGHTERED) '•= (NUMBER OF BEEF AND DAIRY COWS CULLED) +• (NUMBER OF FEEDLOT CALVES AT START OF YEAR) + (NUMBER OF FEEDLOT LONG YEARLINGS AT START OF YEAR) + (NUMBER OF FEEDLOT YEARLINGS ADJUSTED FOR' DEATH LOSS) + (NUMBER OF REPLACEMENTS CULLED) + (NET IMPORT OF SLAUGHTER ANIMALS) (4.g) ABC beef balance (QUANTITY OF < (QUANTITY OF ABC BEEF USED) ABC BEEF SUPPLIED) Where (QUANTITY OF ABC BEEF USED) = (QUANTITY OF DRESSED BEEF SHIPPED OUT OF PROVINCE) + I (NUMBER OF SLAUGHTER CATTLE SHIPPED OUT OF PROVINCE) * (AVERAGE WEIGHT OF SLAUGHTER ANIMAL)! + (QUANTITY REQUIRED TO MEET PROVINCIAL PORTION OF DOMESTIC DEMAND) and (QUANTITY OF ABC BEEF SUPPLIED) = (YIELD FROM FEEDLOT CALVES) +• (YIELD FROM FEEDLOT LONG YEARLINGS) +• (YIELD FROM FEEDLOT YEARLINGS) + (YIELD FROM CULLED REPLACEMENTS) +- (QUANTITY OF DRESSED BEEF SHIPPED INTO PROVINCE) + [ (NUMBER OF SLAUGHTER CATTLE SHIPPED INTO PROVINCE) * (AVERAGE WEIGHT OF SLAUGHTERED ANIMAL)1 (U.h) Pork balance (QUANTITY OF < (QUANTITY OF PORK USED) PORK SUPPLIED) Where (QUANTITY OF PORK USED) = (QUANTITY OF DRESSED PORK SHIPPED OUT OF PROVINCE) + [ (NUMBER OF SLAUGHTER HOGS SHIPPED OUT OF PROVINCE) * (AVERAGE WEIGHT OF SLAUGHTER ANIMAL) ] + (QUANTITY REQUIRED TO MEET PROVINCIAL PORTION OF DOMESTIC DEMAND) and (QUANTITY OF PORK SUPPLIED) = (YIELD FROM HOGS) + (QUANTITY OF DRESSED PORK SHIPPED INTO PROVINCE) +• [ (MUMBER OF SLAUGHTER HOGS SHIPPED INTO PROVINCE) * (AVERAGE WEIGHT OF SLAUGHTERED ANIMAL)] (5) Port orop balance (QUANTITY OF CROP c > (QUANTITY OF CROP e SHIPPED FROM SHIPPED FROM PROVIN- PORT TO PROVINCIAL LEVEL) CIAL LEVEL TO PORT) +• (QUANTITY EXPORTED FROM PORT) (6) National crop balance (QUANTITY OF CROP c X (QUANTITY SOLD IN TRANSFERRED TO NATIONAL NATIONAL MARKET) LEVEL) (7) National livestock balance (7.a) D beef balance (QUANTITY SOLD IN < (CULLED BEEF AND NATIONAL. MARKET) " DAIRY COWS) (7.b) Balance of other livestock products (QUANTITY SOLD IN < (QUANTITY PRODUCED NATIONAL MARKET) ~ WITHIN PROVINCES) ( 8 ) National livestock trading pool balance (QUANTITY SHIPPED < (QUANTITY SHIPPED FROM NATIONAL POOL) INTO NATIONAL POOL) (9) International livestock trade [ (TOTAL EXPORTS) - (TOTAL IMPORTS) ] = (QUANTITY SHIPPED FROM EACH PROVINCE TO WORLD) - (QUANTITY SHIPPED FROM WORLD TO EACH PROVINCE) + (QUANTITY SHIPPED FROM POOL TO WORLD) - (QUANTITY SHIPPED FROM WORLD TO POOL) (10) Objective function (maximand) [ (TOTAL REVENUE) - (TOTAL EXPENSES) ] where (TOTAL REVENUE) = (EXPORT SALES OF CROPS) + (EXPORT SALES OF BEEF PRODUCTS) + (EXPORT SALES OF HOG PRODUCTS) + (DOMESTIC SALES OF CROPS) . + (DOMESTIC SALES OF BEEF) .!:"; + (DOMESTIC SALES OF PORK) + (DOMESTIC SALES OF DAIRY PRODUCTS) + (DOMESTIC SALES OF POULTRY PRODUCTS) and (TOTAL. EXPENSES) = (SUM OF PROVINCIAL PRODUCTION COSTS) + (NATIONAL. TRANSPORTATION COSTS) +• (IMPORT PURCHASES OF BEEF PRODUCTS) + (IMPORT PURCHASES OF HOG PRODUCTS) 63 2.3-8 Summary The purpose of this chapter was to present the theoretical basis for the sectoral model developed for this study and to describe i t s structure. In doing so, reference is made to the overall model framework and to specific submatrices of i t . As with a l l models, alternative methods of representing the structure of a sector may be chosen. The author has emphasized those components of the system that deal with the problem being studied, namely the introduction of a new wheat variety. The simplifying assumptions in the current version of the model w i l l be reviewed in Section 4 .4 , and a number of suggestions w i l l be made in Section 5.3 for broadening the model's scope. However, even in i t s present form, the^ usefulness of the model extends beyond this current analysis. In the chapter that follows the data requirements of the model are noted. 64 CHAPTER 3 The Empirical Model In the last chapter, the general structure of the model was presented. Activities and constraints of the model were described but no coefficient values were noted. The purpose of the present chapter i s to address the empirical content of the model. The data requirements w i l l be discussed i n general terms and the organization of this data and their transformations w i l l be described. This description of the empirical model i s also presented i n Chapter 3 of the associated report entitled "The Structure of CRAM: A Canadian Regional Agricultural Model" (Webber et a l . ) . As noted i n Section 2.3, the model consists of approximately 1180 activ i t i e s and 725 constraints. Given these dimensions, the number of elements, or coefficients, of the model i s potentially very large. Although many elements are zero, the number of non-zero coefficients i s s t i l l considerable (approximately 4900), and consequently the data requirements of the model are extensive. In Section 3.1 an overview of the data i s presented and related to the general structure of the model as presented i n Chapter 2. The organization of this data into sets or f i l e s i s reported on i n Section 3-2, and abstracted portions of these are noted i n Sections 3.3 to 3.7. A complete l i s t i n g of the data and data sources are presented in the Technical Appendix1 of this thesis. A copy of the Technical Appendix i s available from the Department of Agricultural Economics at the University of British Columbia. 65 3.1 General Description of the Data The ac t i v i t i e s i n the model can be divided into production, shipping, and marketing ac t i v i t i e s , and this division was used in describing the general structure of the model (Figure 2.1). The data requirements of each of these major blocks are summarized i n Table 3.1. Table 3.1: Summary of Major Data Requirements of the Model 1. Production sub-matrices: Crop (by region) . land available . crops grown . planted:fallowed ratios . cropping ratios . cost and yield data . % of yield for misc. uses Livestock (by Province) . number of animals i n opening and closing stock categories . diet, cost, and yield data . culling, replacement, and calving rates for cows . death loss rates 2. Shipping sub-matrices: . commodities shipped . shipping routes . unit shipping costs . volumes shipped . average weight of shipped . opening and closing crops 3. Marketing sub-matrices: . commodity sold . commodity prices . domestic demand levels . provincial demand weights slaughter animals inventory level of In general, 1984 was taken as the base year for coefficients in the model. When i t was more desirable to use data pertaining to five year averages, the interval 1980 to 1984 was used. In special cases, the values 66 applied to longer periods. An important example i s the yield data for medium quality wheat which are based on f i e l d t r i a l s dating back to 1974. Many different data sources were used. Most represent published documents from Statistics Canada, Agriculture Canada, and provincial Ministries of Agriculture as noted i n the Technical Appendix where these data are presented. In addition, a deliberate attempt was made to f u l l y u t i l i z e the National Farm Survey results for production data. This well-established survey not only provides a centralized and consistent source of information for the current version of the model, but also simplifies the updating process. Crop regions i n the model correspond with those defined i n the survey. Two general approaches are possible i n developing a mathematical programming model. Fi r s t , the row, column, right hand side, and structural bound sections can be developed directly as coefficients for the model. Second, a special purpose matrix generator can be used to create the matrix automatically. Due to the substantial amount of data involved and the need to test a potentially large number of policy options, the matrix generator approach was selected. The data for the model was f i r s t organized into logical groupings and stored in computer f i l e s . A computer program (the matrix generator) was then developed to read the data from these f i l e s , perform any necessary transformations, and output the model coefficients i n MPSX column format (IBM). The appropriate measurement system (imperial or metric) to be used also had to be determined. Much of the historical data i s i n imperial units. On the other hand, the model output i s required i n metric. The solution was to allow conversion factors to be stored with the data, and then to program the 67 matrix generator to perform any specified transformations. Consequently, the generated coefficients are always i n metric. 3.2 The Fi l e Structure Before describing the data i n detail, an overview of the entire data management system i s provided. Each f i l e i s defined as a set of related data which pertains to a given submatrix or section of the model. Data for the column section of the model i s organized into sixty-two input f i l e s . F i fty-five of these can be classified into seven major types as indicated i n Table 3.2 and the seven remaining f i l e s essentially contain control information or are designed for future model developments. The primary purpose of the four format f i l e s i s to describe the format of the crop, beef, and livestock f i l e s to the matrix generator. Besides the constraint coefficients within the matrix, there are objective function coefficients, right hand side variables, and structural bound sets (Figure 2.1). Corresponding to these three sets of data are three f i l e s : 1) objective function f i l e : commodity prices 2) "right hand side" f i l e : land available by region : opening and closing stock of beef by category 3) structural bounds f i l e : number of hogs, dairy cows, broilers, and egg-layers by province : volumes shipped along certain routes : domestic demand levels for certain commodities 68 Table 3.2: Summary of Input Data Files by Category Production Data 1. crop data f i l e s . cost and yield data 2. crop ratio data f i l e s . crops grown . cropping ratios . "planted:fallowed" ratios 3. beef data f i l e s . diet, cost, and yield data for beef 4. beef coefficient f i l e . culling, replacement, and calving rates for cows 5. livestock data f i l e s . diet, cost, and yield data for hogs, dairy, and poultry Shipping Data 6. shipping data f i l e s . shipping routes by commodity . unit shipping costs by route Marketing Data 7. domestic data f i l e . commodities sold . provincial population weights Miscellaneous Data Number of Files 29 (1 per region) (1 per province) 3 (1 per province) (1 per province) (1 each for crops beef, and hogs) 8. control information f i l e s . western crop format f i l e . eastern crop format f i l e . beef format f i l e . livestock format f i l e . crop inventory f i l e (not used) . canola processing f i l e (not used) . crop shipment coefficient f i l e (not used) _J_ Total Number of Files 62 a For crop production, each Maritime province constitutes a region. For livestock production, the Maritimes are grouped together. 69 The matrix generator i s composed of eight program modules, or subroutines, corresponding to the various blocks of the model. The relationship between the input f i l e s and the generator subroutines i s given i n Figure 3.1. The sequential nature in which these data f i l e s are accessed i s illustrated and the correspondence between each data set and a block of the model i s shown. In the following sections, the data requirements are discussed i n greater detail with reference to representative f i l e s from those summarized in Table 3.2. As noted earlier, a complete l i s t i n g of a l l data i s included i n the Technical Appendix. 3.3 The Crop Production Sector Crop production information i s contained i n crop data f i l e s , crop ratio data f i l e s and crop format f i l e s . Crop data f i l e s specify cost and yield data for crops grown i n the different regions considered, whereas crop ratio data f i l e s present historical records of plantings i n each of these areas. There i s one crop data format f i l e for western Canada and one for eastern Canada. Each contains information on the proportion of the yield to be deducted for seed and wastage by crop, a description of the layout of the crop data f i l e s , and the required conversion factors for the cost and yield data. The crop data f i l e s and crop ratio data f i l e s w i l l be elaborated upon in the following sections. 3.3.1 Cost and Yield Data As noted i n Table 3.2, there i s one crop data f i l e for each of the 29 crop regions where cost and yield data for each crop grown are noted. The crop data f i l e for region 1 of Alberta i s used for i l l u s t r a t i o n (Table 3.3). S22S§ il Se9 - 5 C » 3J fi 0 > ' l 3 E a i l ii n 1 I] 'Si fi < 5 » L i a II S <E» : i <=»• e > H | J l i = M I •( C ** a " « 3 u o a .2 . C s 2 * a a M f S 2 ^ 82 ;ss H E =3 s M > < U •« to A < g • w * -» »• " S 8 fi 5 " .1 ? 5 • R 1: EE §1 > — . - < •» 1 o " ! = * a I 1 ! Sr.*: al . . * 9 1 - -i l l r-1 CD cn (0 rd Q 4J I H 1 (0 8 t5 0 3 fa Table 3.3: Portion of Crop Data for Alberta, Region 1 CROP BLOCK (REGIONAL) PROVINCE: ALBERTA REGION : l PART I HIQ WHEAT S.F. I HIQ I WHEAT I STBL MED WHEAT S.F. I MED I WHEAT I STBL I I I I FERTILIZER I 2.82 I 4.84 I 2.82 I 4.84 I 2.82 I 3.71 CHEMICALS I 1.51 I 1.51 I 1.51 I 1.51 I 1.51 I 1.51 FUEL I 7.71 I 7.71 I 7.75 I 7.75 I 7.10 I 7.10 REPAIR MCH I 6.63 I 6.63 I 6.70 I 6.70 I 6.37 I 6.37 WAGES I I I I OTHER(TOT) I 20.50 I 18.99 I 20.50 I 19.00 I 18.81 I 17.45 -SEED I 10.31 I 9.53 I 10.31 I 9.53 I 8.75 I 8.16 -INSURANCE I 3.64 I 2.89 I 3.64 I 2.89 I 3.64 I 2.89 -REPAIR BLGI .68 I .68 I .68 I .68 I .68 I .68 -UTILITIES I .79 I • .79 I .79 I .79 I .79 I .79 -INTEREST I 1 .87 I ' 1 .89 I 1 .87 I 1.90 I 1 .74 I 1 .72 -MISC I 3.21 I 3.21 I 3.21 I 3.21 I 3.21 I 3.21 YIELD I 30.40 I 20.40 I 35.71 I 23.96 I 43.70 I 38.50 PART 2 I I I I I CANOLA I CANOLA I FLAX I FLAX I OTHER I OTHER I S.F. I STBL I S.F. I STBL I S.F. I STBL I I I I FERTILIZER I 4.08 I I 3.43 I 3.56 I 8.12 CHEMICALS I 6.68 I I 3.48 I 7.08 I 7.08 FUEL I 6.82 I I 9.02 I 9.64 I 9.64 REPAIR MCH I 8.26 I I 7.95 I 15.25 I 15.25 WAGES I I I I OTHER(TOT) I 12.17 I I 18.47 I 60.42 I 60.42 -SEED I 2.04 I I I -INSURANCE I 3.64 I I I -REPAIR BLGI .68 I I I -UTILITIES I .79 I I I -INTEREST I 1.81 I I I -MISC I 3.21 I I I YIELD I 23.30 I I 17.70 I 159.00 I 159.00 PART 3 I I I I I SUMMER I TAME I TAME IUNIMPRVDI I I FALLOW I HAY I PASTURE I LAND I I BARLEY I S.F. I BARLEY STBL FERTILIZER CHEMICALS FUEL REPAIR MCH WAGES OTHER(TOT) -SEED -INSURANCE -REPAIR BLGI -UTILITIES -INTEREST -MISC YIELD I I I 3.30 I 2.39 I I I I I I I I I I 7.37 .84 1.91 I I I I I I I I I I I I I ,26 I 0.10 -Units: •the costs for a l l cropping a c t i v i t i e s , from f e r t i l i z e r through to misc, are S/acre •tame hay 6 pasture yields are tons/acre, other crop yields are S/acre and the rest are bu/acre 72 Each column of Table 3.3 refers to a particular use of land i n the region, and the rows refer to the various inputs required as well as the resulting yields. Six f i e l d crops are li s t e d in the table. Furthermore, a distinction i s made between planting on land previously fallowed, and planting on stubble. There are four land types identified i n the model: crop land, tame hay, tame pasture, and unimproved land. The crops l i s t e d i n columns 1 to 13 constitute uses of cropland, while columns 14, 15, and 16, relate to the uses of the remaining three land types. Note that the tame hay category includes greenfeed. The amount of each land type available i n a region i s specified i n the right hand side section of the MPSX f i l e . The set of crops specified i n Table 3.3 i s the same for a l l 23 western crop regions. The set defined for eastern regions i s slightly different. Flax i s omitted, while corn grain, corn silage, and soybeans are added. In addition, "other eastern" crops are distinguished from "other western" crops, and the distinction between planting on stubble and planting on fallow i s not made. The f i r s t six rows of Table 3.3 specify the cash costs ($ per acre) required for each of the cropping a c t i v i t i e s . These are summed to give the total cost. Rows 7 to 12 give the components of the miscellaneous cost item specified in row 6. From a record keeping aspect i t i s convenient to distinguish these costs so that updates are faci l i t a t e d . The f i n a l row specifies the yield. The units of these values depend on the crop activity and are specified at the end of each table. It i s convenient to record some yields in bushels, others i n tons. Recall that the required factors to convert these values to metric are specified i n the crop data format f i l e . The yield adjustment factors to account for seed use and waste are also 73 given i n that f i l e . The crop data f i l e s for the 23 western crop regions contain cost and yield estimates for medium quality wheat. Costs of growing the new variety were considered to be not much different from those associated with other wheats or feed grains. An allowance was made for extra fuel, repair and maintenance costs associated with higher yields. Additional costs were derived with the help of a dryland crops simulation model (Zentner et a l ) . Yields for HY320 (medium quality) and Neepawa (high quality wheat) were obtained from co-operative variety t r i a l s undertaken during the time period 1974-85 by plant scientists i n various locations i n the prairie provinces. The relative yield advantage of medium quality wheat in each t r i a l was assumed to apply i n regions having similar s o i l and climatic conditions (Table 3 . 4 ) . This yield advantage was used to estimate farm level yields of medium quality wheat i n each region, for which no comprehensive data series yet exists. Yields for a l l crops except medium quality wheat are from National Farm Survey data. A summary of cost and yield data for a l l western producing regions i s presented i n Appendices A and B. 3.3.2 Crop Rotations Crop rotation data f i l e s are used to specify the proportion of cropland fallowed, and the set of valid crop mixes for each crop region. The data for region 1 of Alberta i s presented in Table 3 . 5 . The variation in cropping patterns for this region i s considerably less than that observed for some other regions. Note that a complete l i s t i n g of crop mixes i s provided i n the Technical Appendix. Summerfallow acreages vary substantially due to economic and biological considerations. The minimum quantity of summerfallow permitted i n each of Table 3 .4: Relative Yield Advantage3 of Medium Quality Wheat over Hard Wheat ( % Difference) Region Co-op T r i a l Locat ion 1974 1976 Man - 1 Indian Head 33 34 Man - 2 Dauphin & Watrous 27 Man - 3 Portage La P r a i r i e Man - 4 Carman,Indian Head & Winnipeg Man - 5 TeuIon 48 50 Man - 6 Winnipeg 21 28 34 1977 1978 1981 1982 1983 1984 1985 Mean St .Deviat ion 28 7 46 22 26 18 42 23.4 12.2 84 27 38 17 24 36.3 24.1 21 16 33 20 33 24.5 7.6 48 13 26 30 19 18 42 28.9 11.5 37 35 16 - 5 30.2 21.2 67 20 6 43 13 28.3 22.7 Sask - 1 Indian Head 33 34 28 7 46 22 26 18 42 28.4 12.2 Sask - 2 Regina 2 26 37 52 13 7 33 6 41 24.2 17.6 Sask - 3 Stewart Va l ley 22 - 4 38 32 69 68 5 28 25 31.3 24.7 Sask - 4 Swift Current 17 49 21 35 2 34 5 6 -10 17.5 18.9 Sask - 5 Watrous 84 27 38 17 24 38.2 26.5 Sask - 6 Saskatoon 52 55 55 20 51 35 44.6 14.5 Sask - 7 Scott 4 23 17 1 31 15.2 12.5 Sask - 8 Me l for t 24 39 78 23 26 4 18 45 32.2 22.2 Sask - 9 Mel f o r t & Beaverlodge 5 30 32 50 23 26 4 18 45 25.9 15.8 A l ta - 1 Swift Current 17 49 21 35 2 34 5 6 -10 17.5 18.9 A l ta - 2 Acme 16 26 18 22 18 11 18.5 5.1 A l ta - 3 Acme 16 26 18 22 18 11 18.5 5.1 A l ta - 4 Lacombe & Mel for t 28 29 39 78 23 26 4 18 45 32.3 20.7 A l ta - 5 Lacombe & Mel for t 28 29 39 78 23 26 4 18 45 32.3 20.7 A l t a - 6 Beaverlodge & F t . V e r m i l l i o n 14 35 32 32 25 27.8 8.3 A l ta - 7 Beaverlodge & F t . V e r m i l l i o n 14 35 32 32 25 27.8 8.3 B.C. Relat ive Advantage - "Y 32° fiff-" flST ^  * 1001 75 Table 3-5: Proportion of Cropland Planted to Various Crops (%) for Region 1 of Alberta (1979-1984) Period Wheat Barley Flax Canola Other 1981 0.78 0.15 0.01 0.00 0.06 1982 0.77 0.15 0.01 0.01 0.06 1983 0.79 0.14 0.00 0.01 0.06 1984 0.78 0.14 0.01 0.01 0.06 1979-1984 0.76 0.16 0.01 0.01 0.06 the twenty-three western crop regions i s that which occurred i n 1984. During that year, prairie farmers fallowed fewer hectares than i n any year since 1970, the year of the Lower Inventories For Tomorrow (LIFT) program (Klein, et a l ) . For region 1 of Alberta, this amounted to 44$ of the cropland available. For each region, five crop mixes are specified based on historical observations between 1979 and 1984. For a given set of product prices, the most profitable crop mix satisfying the regional constraints i s selected. This approach was proposed by McCarl to address the aggregation problems associated with regional models. McCarl argues that the main advantage of this approach " i s the derivation of sensible crop mixes i n the microeconomic context and the application of this advantage at the regional level where the model w i l l be solved more easily without the use of f l e x i b i l i t y constraints." 76 3.4 The Livestock Sector Five f i l e types are used to describe livestock data including three for beef information, and two for the other livestock categories (hogs, dairy cows, and poultry). They are the beef data f i l e s , the beef coefficient f i l e , livestock data f i l e s , beef data format f i l e and the livestock data format f i l e . The f i r s t three types contain diet, cost, and yield information, as well as certain biological and herd management parameters. The last two f i l e s describe the layout of the beef and livestock data f i l e s , respectively, and specify the necessary factors for converting units from imperial to metric. In addition, the beef data format f i l e contains information regarding the distribution of diet requirements over a year. These items are explained more f u l l y below. 3.4.1 Diet, Cost and Yield Data for Beef Production Activities Diets, cost of production and yield information for the beef sector are stored i n the beef data f i l e s , and there i s one such f i l e for each producing area, or province. The data for Alberta i s given i n Table 3.6. Referring to that table, each column corresponds to a particular activity, namely, producing a particular category of beef under a specific diet regime. The f i r s t four rows specify the inputs, and the last two rows specify the yields to be expected from each animal in a particular category. There are seven categories of beef, each associated with at least one diet. The categories are: cows with calf (with three possible diets); cow replacements (one diet); feedlot calves (four diets); feedlot yearlings (four diets); stockers (one diet); pasture yearlings (one diet); and pasture long yearlings (one diet). The f i r s t three rows of Table 3.6 describe the possible diets, i.e. the amount of pasture, forage, and barley required to Table 3.6: Portion of Beef Data for Alberta B E E F BLOCK ( P R O V I N C I A L ) P R O V I N C E : ALBERTA I I I I I F E E D C A F I F E E D C A F PART I I C O W - C A L F I C O W - C A L F I C O W - C A L F I I HI GAIN I HI GAIN : I D I E T I I D I E T 2 I D I E T 3 I R E P L A C E I GRAIN I FORAGE r ! 1 1 j j „ _ I I I I I I PASTURE I 2 . 6 2 3 I 2 .581 I 2 . 5 0 5 I 2 . 8 3 4 I I FORAGE I 1 .401 I 1 .280 I 1 .051 I 1 . 9 3 0 I I 0 .661 : BARLEY I 9 . 9 2 I 1 4 . 2 9 I 2 0 . 4 6 I 8 0 . 4 1 "I 7 4 . 0 0 I 56 .21 CASH COSTS I 2 5 . 0 0 I 2 5 . 0 0 I 2 5 . 0 0 I 1 3 . 0 0 I 1 5 . 6 0 I 1 5 . 6 0 ABC B E E F I I I I S 3 3 . 8 1 5 4 5 . S I 5 4 S . 9 D BEEF 1 4 0 7 . 3 1 4 0 7 . 3 1 4 0 7 . 3 I I I I F E E D C A F I F E E D C A F I F E E D Y R L I F E E D Y R L I F E E D Y R L I F E E D YRL PART 2 ILO GAIN ILO GAIN IHI GAIN IHI GAIN I L O GAIN ILO GAIN I GRAIN I FORAGE I GRAIN I FORAGE I GRAIN I FORAGE j . — _ 1 1 -I 1 I I I I I I PASTURE I I I I I I FORAGE I I 1 .090 I I 0 . 5 3 5 I I 0 . 9 5 2 BARLEY I 8 1 . 4 2 I 5 2 . 1 8 I 7 1 . 9 2 I 5 7 . 5 4 I 7 9 . 9 2 I 5 3 . 4 3 CASH COSTS I 1 5 . 6 0 I 1 5 . 6 0 I 11 . 4 0 I 1 1 . 4 0 I 0 . 2 0 I 0 . 2 0 ABC B E E F 1 546 .1 1 5 4 6 . I 1 6 1 0 . 2 1 6 1 0 . 2 1 6 1 0 . 0 1 6 1 0 . 0 D B E E F I I I I I I I I I F E E D L O T I I I PART 3 I I PASTURE I LONG I I I ISTOCKER I Y E A R L I N G I Y E A R L I N G I I I 1 1 1 1 1 1 I I I I I I PASTURE I I - I I I I FORAGE I 0 . 6 5 5 I 0 . 0 9 3 I 0 . 0 9 3 I I I BARLEY I 11 . 7 5 I 3 3 . 1 2 I 3 3 . 1 2 I I I CASH COSTS I 6 . 1 5 I 1 3 . 0 0 I 0 . 2 0 I I I ABC B E E F I I 1 6 1 0 . 2 1 I I D B E E F I I I I I I - U n i t s : ' p a s t u r e t f o r a g e r e q u i r e m e n t s a r e t o n s / a n i m a l • b a r l e y r e q u i r e m e n t s a r e b u s h e l s / a n i m a l * c a s h c o s t s a r e $ / a n i m a l * a b c & d b e e f y i e l d s a r e l b s / a n i m a l 78 feed one animal in a particular category over one year. Five of the categories above span a calendar year (see Section 2 .3 .2 ) . For these categories the maximum opening stock and minimum closing stock levels are specified for each province i n the right hand side section. In addition, distribution weights are specified for every diet associated with the five categories. For example, suppose a calf enters the stocker category at the start of November, and then i s shipped to the feedlot at the end of April. In this case, one-third of the annual diet i s consumed by a stocker in the closing group, and two-thirds by a stocker i n the opening herd. The weights used to s p l i t a diet between the opening and closing stock are specified i n the beef data format f i l e . Referring once again to Table 3 .6, the fourth row specifies the cash costs incurred, and the last two rows specify the yields. The product i s either high quality beef (ABC grade), or i n the case of culled cows, D beef. Animals transfer from one category to another as they pass through a segment of the beef cycle during the one year period modelled. As the complete cycle i s longer than one year, some animal categories (e.g. stocker and pasture yearling) have no yield associated with them. The beef coefficient f i l e (the second f i l e noted earlier) contains the culling rate, replacement ratio, and calving rate for each provincial beef herd. The values used i n the model are specified i n Table 3 .7 . The culling rate represents that proportion of the beef cow herd culled annually, the replacement ratio i s used to determine the number of beef heifers reserved as replacements to the breeding stock, and the calving rate determines the number of calves entering the system. A l l values are expressed as proportions of the opening cow herd. Another important parameter i s the death rate. It varies with beef 79 Table 3.7: Cull Rates, Replacement Ratios and Calving Rates for the Beef Herd by Province Province Culling Rate Replacement Ratio Calving Rate British Columbia 0.15 0.195 0.72 Alberta 0.09 0.135 0.87 Saskatchewan 0.115 0.160 0.83 Manitoba 0.15 0.195 0.84 Ontario 0.15 0.195 0.70 Quebec 0.15 0.195 0.76 Maritimes 0.15 0.195 0.69 category, but not between provinces. Death rates used are specified as constants i n the model. A value of ^ .5% i s used for cows, replacements, and stookers, 2% for feedlot yearlings, and 3% for feedlot calves, and feedlot long yearlings. 3.4.2 Production Information for Hog. Dairy, and Poultry Sectors Input and yield data required for the the hog, cow, broiler, and egg-layer sectors are contained i n the provincial livestock data f i l e s . The data for Alberta i s presented i n Table 3 .8 . The f i r s t row specifies the number of animals i n each group, and these values w i l l appear as structural bounds i n the model. The second row summarizes the cash costs per animal unit. The next four rows describe the livestock diets expressed as the amount of forage, pasture, wheat, or barley required to support one animal for one year. The following three rows give the culling rate, replacement ratio, and calving rate for the dairy herd (analogous to the beef coefficients presented i n Table 3 .7 ) . Finally, the last section specifies the yield for each animal unit. Dairy cows produce 80 Table 3.8: Livestock Coefficients Required for Hog, Dairy and Poultry Sectors of Model - Alberta Livestock Category Item Units Hog Dairy Broilers Layers Animals No. 1,679,600 203,600 32,562,000 2,318,000 Total Costs $/Au. 17.56 478.15 0.19 2.11 Dietary Requirements Forage tons/Au. 1.594 Pasture tons/Au. 0.797 Wheat bu/Au. 0.12 1.07 Barley bu/Au. 24.80 23.88 " Dairy Coefficients Cull rate ratio 0.15 Replacement rate ratio 0.29 Calving rate ratio 0.87 Yield Pork (cold) lb/hog 165.4 Milk litres/cow 2,778.2 Veal - animals No. 20,700 - cold weight lb/calf 196.6 D-beef (cold) lb/Au. 407.3 Poultry lb/bird 3.73 Eggs dozens/layer 16.01 three items: milk, veal, and D beef. In addition, some dairy calves are transferred to the beef feedlot where they w i l l subsequently yield ABC beef. 3.5 Shipping Routes and Costs Transportation a c t i v i t i e s are specified for certain crops, beef cate-gories, and hog products. There are three f i l e s corresponding to these groups, and each f i l e contains a description of shipping routes and associ-ated rates. The crop shipment f i l e w i l l be discussed here because of i t s importance to the current study, but a l l three f i l e s have the same basic format, as indicated i n the Technical Appendix of this report. 81 The model provides considerable f l e x i b i l i t y in defining shipping a c t i v i t i e s . The essential information required i s point of origin, destination, the commodity being moved and the unit shipping cost. In the case of crops, the origin and destination can be a region, a province, a port, or, for international trade, the world. The commodity moved can be any of the grains included i n the model. However, special attention i s given to the transportation of wheat (both qualities), barley, flax, canola, and corn. The general structure of current crop shipping routes has regional production f i r s t transferred to the provincial level where i t can be used for provincial needs, which include livestock feed and domestic demand. The residual can then be shipped to defi c i t provinces, or to port for export. This approach circumvents the problem of dividing a region's yield between intra- and extra-provincial uses. Currently, two ports, Vancouver and Thunder Bay, are specified. Alberta exports through Vancouver, Manitoba ships through Thunder Bay, and Saskatchewan ships through both ports. As shipping rates are c r i t i c a l to the current analysis, the region-to-port rates are li s t e d i n Table 3.9. A l l rates are expressed as $/tonne of grain regardless of grain type. Note that the impact of a change i n the rate structure can be easily analyzed by modifying the rates specified i n the shipping f i l e . A l l grain transportation routes are decomposed into two segments which include a region-to-province transfer and province-to-port shipment. Given the region-to-port rates for a l l crop regions i n a province, the lowest rate i s assigned to the province-to-port route. For each region, the rate assigned to the region-to-province transfer i s the difference between the province-to-port rate and the region-to-port rate. For example, the rate from region 1 in Alberta to Vancouver, $7.82 per tonne, can be decomposed 82 Table 3.9: Shipping Rates for Grain from Various Prairie Points to Export at Vancouver or Thunder Bay Province Region Reference Centre Rate to Vancouver Rate to Thunder Bay Alberta Saskatchewan Manitoba ($/tonne) ($/tonne) 1 Medicine Hat 7.82 2 Vulcan 7.31 3 Olds 7.19 4 Viking 7.31 5 Red Deer 7.44 6 Barrhead 7.19 7 Falher 7.82 1 Grenfell 10.03 6.55 2 Weyburn 9.84 6.81 3 Moose Jaw 9.09 7.19 4 Fox Valley 8.46 8.58 5 Foam Lake 10.62 6.93 6 Saskatoon 8.58 7.57 7 Rosetown 8.96 7.95 8 Melfort 9.25 7.57 9 Prince Alberta 8.84 7.82 1 Brandon 5.92 2 Grandview 6.30 3 Swan River 6.68 4 Elm Creek 5.41 5 Winnipeg 5.16 6 Arborg 5.54 Source: Canadian Freight Association into $0.63 per tonne for the region to province transfer, and $7.19 per tonne for the Alberta to Vancouver shipment. This method of calculating the cost of region to province transfers lead3 to two possible rates for each Saskatchewan region depending on whether the destination i s Vancouver or Thunder Bay. To resolve the problem,- a region i n western Saskatchewan can either ship to Thunder Bay via the Saskatchewan provincial level, or to Vancouver via the Alberta provincial level. Thus, these regions can supply provincial requirements of both provinces. Likewise, regions i n the east of 83 Saskatchewan can ship to Vancouver via the Saskatchewan provincial level, or to Thunder Bay via the Manitoba provincial level. Finally, i f the unit shipping costs have been expressed in imperial units, conversion factors can be specified at the start of the shipping f i l e . For certain shipping a c t i v i t i e s , upper and lower bounds on the quantity of goods shipped can be specified. In these cases, the limits are given i n the structural bounds section. 3.6 The Domestic and Export Market Sectors Marketing data requirements include quantities sold domestically, and prices received. For convenience, i t was decided that demand levels and domestic prices would not be endogenized at this stage (method as per Duloy and Norton). The relationship between domestic and export demand i s noted in this section, along with a discussion of commodity prices used i n this study. 3.6.1 Domestic Requirements versus Export Levels Production of the agriculture sector i s used locally for feed, processing, or household consumption. Once national requirements are met, the residual can be exported. Domestic demand i s specified for western Canada, for eastern Canada, and, for some commodities, at the national level. The commodities designated at the western or eastern Canadian level are identified i n the domestic demand data f i l e . Currently, they are high quality wheat, barley, ABC beef, and pork. Demand levels for these products are specified i n the structural bounds section, and these requirements are distributed across provinces based on population. The provincial population weights for western and eastern Canada are also stored in the domestic demand f i l e . 84 Demand levels are not set for most commodities sold on the national market and consequently no attempt i s made to distribute their demand over the provinces. As a l l production in these goods can be sold outright, they w i l l not be traded or exported. The one exception i s canola. Its domestic demand has been specified at the national level. Grain requirements for livestock (beef, dairy, hogs, and poultry) are specified i n the beef and livestock data f i l e s where the annual diets are expressed per animal unit (see section 3.4). For beef and dairy, the diets are specified i n terms of barley, forage, and pasture, for hogs i n terms of barley, and for poultry i n terms of medium quality wheat. Some substitution among grains for livestock feed use i s permitted based on their respective digestible energy contents. Allowable substitutions inolude high quality wheat for medium quality wheat, medium quality wheat for barley, corn grain for barley (eastern Canada), and stored forage for pasture. The size of each provincial beef herd i s specified i n the right hand side section, while the numbers of animals i n the dairy, hog, and poultry categories are defined i n the structural bounds section. These totals, along with the actual diets selected, determine the amount of feed grain required. In the case of medium quality wheat production, the amount available for export can be determined once livestock feed requirements have been met. 3.6.2 Commodity Prices The prices of a l l crops i n the model (excluding medium quality wheat) were arbi t r a r i l y set at their 1984-85 level based on Thunder Bay (Table 3.10). 85 Table 3.10: Grain Prices Selected for the Base Run - 1984-85 Average Prices Crop Price ($/tonne) High quality wheat 186 Barley 131 Flax 355 Canola 383 The price of medium quality wheat i s somewhat d i f f i c u l t to gauge, since very small quantities of this product have been sold by the Canadian Wheat Board. In 1984-85, the f i n a l realized producer price for HT320 was $170/tonne, just slightly lower than the price for the 3CW grade of hard wheat. However, this price was determined i n the domestic market, whereas the export price i s desired. Ulrich and Furtan i n their analysis assumed the price would f a l l somewhere between the price of 3CW and Canada Feed grades. The f i n a l realized price for Canada Feed wheat i n 1984-85 was $141.50 per tonne. Although livestock feed would represent the lowest value use of HT320, the effect on price of an increase i n supply of feed wheat i s not known. Concerning the world market, most analysts agree that Canadian produced medium quality wheat varieties would be priced between Australian soft white wheat and American medium-hard wheats at major destinations (Gilmour). The 1984-85 prices of these wheats are compared with that of no. 1 CWRS in Table 3.11. The comparable values for Australian soft wheat and U.S. hard winter and U.S. western white wheat during the period 1980 to 1985 were 0.77, 0.79, and 0.72, respectively. Thus, i t would appear the price of HY320 could go as low as $135 per tonne under the "small country" assumption that Canada w i l l have no appreciable effect on world price. 86 Table 3-11: The 1984-85 Price for No. 1 C¥RS and the Relevant Price Range for Triple-M Wheat Varieties ($ Canadian) CIF Japan Inferred Transport Farm Gate Relative Grade Price & Transaction Costs Price Price No. 1 CWRS 284 98 186 1.00 ASW 237 98 139 0.75 USHW 241 98 143 0.77 OSWW 223 98 125 0.67 CWRS - Canadian Western Red Spring Wheat ASW - Australian Soft Wheat USHW - United States Ord. No. 2 Hard Winter Wheat USWW - United States No. 2 Western White Wheat Considering the uncertainty regarding the ultimate price of Canadian medium quality wheat, as well as the importance of that price to the results, i t was decided to perform the current analysis over a range of possible prices. The set used was: $135, $140, $145, $150, $155, $160, $165, and $170 per tonne. 3.7 Right Hand Sides and Structural Bounds Upper and lower bounds can be placed on constraints or on ac t i v i t i e s of a model. In the f i r s t case, they are implemented as right hand sides, and in the second instance, they are specified as structural bounds. In the current model, the non-zero right hand sides consist of land available by type, and opening and closing stocks of beef by category. Four different types of land classes are identified, namely, crop land, tame hay, tame pasture and improved land. For the 29 crop regions included in the model, an upper lim i t on each of these four land classes i s specified. Currently, a transfer of land from one class to another i s not possible although i t i s quite feasible to specify act i v i t i e s which allow 87 such conversion. An upper bound on the opening stock and a lower li m i t for the closing stock are specified for five categories of beef: beef cows, replacements, stockers, feedlot calves and feedlot yearlings. These bounds are specified for each of the seven provincial beef herds. The implication i s that herd size i s exogeneous and that the flow between animal classes during the production year i s controlled. However, the opening and closing stock numbers can be varied between runs so that various scenarios can be evaluated. As mentioned i n Section 3.1, structural bounds are specified for the number of hogs, dairy cows, broilers, and egg-layers by province, for the volume of commodities shipped along certain routes, and for the domestic demand levels of certain commodities. Complete l i s t i n g s of the right hand side and structural bound f i l e s are given i n the Technical Appendix. In summary, this chapter has presented information on the coefficients of the model. The approach followed has been to provide selected examples of these data with a complete l i s t i n g thereof presented in the Technical Appendix. The organization of these values into sets allows for easy modification of a l l coefficients. This f l e x i b i l i t y i s important i f the model i s to be used to analyze other policy issues. Other questions that have been examined to date include alternatives to the "Crow11 freight rate structure, the impact of the 1985 U.S. farm b i l l on Canadian agriculture, and the possible impacts of a national marketing scheme for hog production (Klein et a l . , MacGregor and Graham). 88 CHAPTER 4 Testa and Results This chapter discusses the tests conducted u t i l i z i n g the sectoral model presented i n Chapters 2 and 3, and reports on the results obtained. There are four sections: Section 4.1 briefly reviews problems i n model validation and outlines the procedure used to validate the current model; Section 4.2 describes the parameter values used to test the economic impact of introducing medium quality wheat on the prairies; Section 4.3 analyzes the results; and, f i n a l l y , Section 4.4 reviews the simplifying assumptions of the model that are relevant to the interpretation of the results. 4.1 Model Validation It has been noted by McCarl that true model validation can rarely be achieved, but that i t i s possible by systematic procedures to improve the relevancy of models. He argues that na model need not mirror the perceived reality perfectly; rather, i t needs to abstract reality adequately for the model's anticipated use." Three general uses are recognized: structural exploration, prediction, and prescription. The current model i s designed to be predictive, in particular, to forecast the consequences of introducing a new variety of wheat into production. In this setting, validation indicates the degree to which the predictions should be believed. McCarl outlines seven categories of validation tests i n increasing order of complexity. They are: 89 1. Plausibility Test 2. Possibility Test 3. Supply Function Test 4. Dual Supply Function Test -5. Prediction Test 6. Predictive Change Test 7. Predictive Tracking Test Does the model create "plausible" results? Does the model duplicate a "reality" situation? Is the marginal cost of production close to the observed price? With prices fixed at expected prices does output compare with actual output? Is the model able to predict out-comes when specified with parameters identical to those leading to that outcome? Models may not need to predict exactly as long as they predict the magnitude or possibly even the direction of change accurately. In this test the a b i l i t y to predict a one time change i 3 not that important as compared to the a b i l i t y of the model to predict change over time. . Each test would require a set of data consisting of parameters with an associated set of logical outcomes. The outcome set could then be compared with the results obtained by running the model with the test parameters as input. If the two sets were sufficiently similar, based on some association measure, then the test would be successful. Since the base data for a model 90 often represents "typical" values rather than "actual" values for a particular time period, there remains a problem with the basis for comparison of the model results. Clearly compromises have to be made. McCarl outlines a six step iterative procedure for applying the tests, or a subset thereof, in increasing order of complexity (Table 4.1). Although this process should improve the relevancy of a model, the ultimate test Is whether or not i t i s used for i t s intended purpose. As i s readily apparent from this description, model validation involves a considerable degree of subjectivity. In the process of testing and validating the current model, several of the technical tests were performed according to the procedure suggested by McCarl. The results are summarized below, whereas a more complete discussion i s contained i n Webber et a l . (pp. 73-102). The i n i t i a l specification of the model resulted i n i n f e a s i b i l i t i e s which indicated misspecifications of the structure. Following a re-examination of the concepts involved, modifications were made. Particular d i f f i c u l t i e s were encountered i n attempting to validate the beef sector of the model on a province by province basis, partly due to inconsistencies i n various data sources. In addition, the classification of animals by the various age categories necessitates assumptions regarding birth and death rates, as well as the aging of animals from one category to another. This area requires further work. Steps 2 and 3 of the procedure involve solving the model for particular levels of the decision variables, and evaluating the results. In applying the "possibility test", questions regarding the a b i l i t y of the model to duplicate the 1984 base were examined. When results were inconsistent with expectations, corrections were made, and steps 1 and 2 of the validity 91 Table 4.1: A Procedure For Model Validation Step 1. Enter the parameters associated with a particular validation test along with any special constraints or ac t i v i t i e s . Step 2 . Attempt to solve the model. Step 3. Evaluate the results. There are 2 p o s s i b i l i t i e s : (a) I f the model has failed to solve, search for a deficiency in the structure. If possible, correct the problem and go to step 2 ; otherwise, go to step 6. (b) If the model has a solution, use association measures to compare the outcome set with the model solution. If the measures indicate a sufficient degree of association, go to step 4; otherwise, go to step 5. Step 4. If satisfied that the model i s "not invalid", accept the model for use; otherwise, prepare to do a more complex test and go to step 1. Step 5. Search for a deficiency in the test data, the input data, or the model. If possible, correct the problem and go to step 2; otherwise, go to step 6. Step 6. Evaluate the model. There are 3 p o s s i b i l i t i e s : (a) If the model should be revised, make the necessary revisions. If the revisions are not extensive, go to step 2; otherwise, prepare to do a less complex step and go to step 1. (b) If the model should be qualified, select any of: (i) If the current test i s incomplete, go to step 2 . ( i i ) Prepare to do a more complex test and go to step 1. ( i i i ) Accept the model for use with qualification. (c) I f the model should be discarded, terminate the procedure. procedure repeated. This testing involved approximately 50 different solutions. The outcome set consisted largely of data available from the Canada Grains Council which detail annual flows of the major grains for each of the prairie provinces. Although these flows are averaged over the five year 92 period 1976 to 1981, they demonstrate typical movements and use of grains within western Canada, and i t i s this aspect of Canadian agriculture that has received the attention of this study. Further information on some of these values were provided by the Canada Grains Council publication "Sta t i s t i c a l Handbook 1984". Turning to results of the model, 1984 was used as the base year where possible, although part of the input series involved annual data over the five year period 1979 to 1984. Base case results are presented i n Tables 1 to 25 i n the Technical Appendix. The levels of a l l a c t i v i t i e s i n the model are reported, and i n addition, penalty costs are included for those ac t i v i t i e s entering with a zero value. The information i s organized on a province by province basis with livestock production being documented f i r s t , followed by the cropping a c t i v i t i e s by region. The crop, beef, and hog shipping a c t i v i t i e s are then presented and, f i n a l l y , the domestic demand levels are noted. In the predictive change test where the magnitude or even the direction of change needs to be predicted i t i s considered that the model has been validated. In terms of some of the other technical validation exercises, one may debate the degree to which the present model has f u l f i l l e d a l l test c r i t e r i a . For example, in the dual supply function test where production at an observed price level i s tested against observed production, with a l l constraints including crop ratio constraints removed, i t may be questioned whether this test has been adequately passed. As with a l l research models there are deficiencies that may be identified, but these can be addressed with time. Meanwhile, there are many interesting problems to which the analytical framework can be usefully applied including the issue of regarding medium quality wheat. 93 4.2 Incorporating Medium Quality Wheat The objective of the thesis i s to determine the impact on the western provinces of licensing medium quality wheat. More specifically, i t i s to assess the optimal level and location of grain production that maximizes net returns to the agriculture sector for various relative prices of medium quality wheat. Furthermore, the effects on export levels and revenue must be determined. The rate of adoption of the new variety w i l l be related to the relative costs, yields, and prices of the major alternative crops. These values w i l l be summarized i n this section. Data on comparative costs and yields of HY320 (the variety of medium quality wheat recently licensed) were assembled for each of the 23 prairie crop regions. Costs of growing the new. variety were considered to be similar to those connected, with growing other wheats or feed grains, although an allowance was made for expenses associated with higher yields. Appendix A contains the variable production costs associated with each crop by prairie region. The relative yield advantage of HY320 over Nepawa (the dominant traditional variety) for each crop region was derived from co-operative variety t r i a l s undertaken i n various locations on the prairies during the period 1974 to 1985. Although the results varied by region and by whether i t was grown on fallow or on stubble, on average HY320 outperformed Nepawa by 27 percent (Table 3.4). These relative advantages were applied to documented Nepawa yields to derive estimates for medium quality wheat for each of the prairie crop regions. Regional yields for a l l other crops were obtained from National Farm Survey data. The crop yields are list e d by prairie region i n Appendix B. The land available for each crop within a region was constrained by 94 crop ratios defined i n the model, as well as by the constraint on the quantity of land fallowed. The crop ratios are based on data for the five year period between 1980 and 1985, and reflect rotational constraints. The summer fallow constraints specify lower bounds on the amount of land fallowed i n the 23 prairie regions. These minimum values are based on the levels which occurred i n 1984. Since data were not available on many potential constraints on medium quality wheat production, i t was i n i t i a l l y assumed that the new variety could only compete directly with high quality wheat. This assumption was implicit i n the studies of Henning, Ulrich and Furtan, and Loyns and Carter. Indirect substitution of medium quality wheat for other grains was possible through a change i n the optimal crop mix and fallow ratio. This constraint on the quantity of cropland available for the new wheat variety was later relaxed to allow medium quality wheat, hard wheat, and barley to compete for land jointly allocated to wheat and feed grains. In addition to the constraints on medium quality wheat acreage described above, the degree of substitution was further restricted by the necessity of meeting domestic demands of traditional varieties for human consumption and industrial use. Furthermore, livestock feed requirements had to be satisfied, although some substitution among grains for this use was permitted. In particular, high quality wheat could substitute for medium quality wheat, and medium quality wheat for barley. Appropriate conversion factors were applied. Prices of traditional crops were set at their 1984-85 level based on Thunder Bay. These prices are: $l86/tonne for high quality wheat, $131/ tonne for barley, $355/tonne for flax, and $383/tonne for canola. The price realized at the regional (or "farmgate") level was determined by subtracting / 95 the appropriate shipping rate from the Thunder Bay price. The regional shipping rates for western grain are li s t e d i n Table 3 .9 . The price of medium quality wheat i s d i f f i c u l t to predict since very l i t t l e has been sold to date by the Canadian Wheat Board. As noted in Section 3.6.2, the estimates range from $135 per tonne (the lower range for American and Australian medium quality wheat currently traded on the world market) to $170 per tonne (the f i n a l realized price for HY320 i n 1984-85). Due to the uncertainty concerning the ultimate price, as well as i t s importance to this study, the analysis was conducted at eight specific prices over the range $135/tonne to $170/tonne (Table 4.2). Table 4.2: Parametric Variations i n the Price of Medium Quality Wheat (MQW) Level Price of MQW Relative Prive of MQW3 ($/tonne) 1 135 0.72 2 140 0.75 3 145 0.78 4 150 0.80 5 155 0.83 6 160 0.86 7 165 0.88 8 170 0.91 a Price of medium quality wheat divided by price of high quality wheat ($l86/tonne). 4.3 Results The model was solved with medium quality wheat excluded i n order to obtain a base case. The resulting aggregate production levels are discussed i n Section 4.3.1. Next, medium quality wheat was introduced by allowing i t 96 to compete with hard wheat for land allocated to wheat production. Eight separate runs were conducted corresponding to different selling prices of medium quality wheat. The output of these price scenarios was compared with that of the base case to evaluate the impact of introducing the new variety on production, exports, and income. The findings are reported i n Sections 4.3.2, 4.3.3, and 4.3.4 respectively. Finally, in Section 4.3.5, the results of relaxing the restriction on land available to medium quality wheat are reported. In this section, the new variety i s allowed to compete with both hard wheat and feed grains (represented by barley) for land jointly allocated to those crops. Because the model generates a large amount of detail, the following aggregated format w i l l be used to present the results. The term "feed grains" w i l l refer to barley and barley equivalents (oats and mixed grains); canola and flax w i l l be summed under the category "oilseeds"; and the distinction between "grown on fallow" and "grown on stubble" w i l l not be shown. Furthermore, the results for the category "other crops" w i l l normally be suppressed. 4.3.1 Base Case The base case i s that solution of the model in which medium quality wheat i s not available as a crop alternative. The results for prairie grain production levels are presented i n Table 4.3. These represent an optimum planting scheme within the bounds of the various crop rotation a c t i v i t i e s . The solution to the model specifies the acreage allocated to each crop by crop region, and these levels have been aggregated up to the provincial and prairie level as indicated. 97 Table 4.3: Model Results for the Base Situation - Area Planted to Each Crop by Western Province (Thousand Hectares) Grain B.C. Alberta Saskatchewan Manitoba Total Hard Wheat 74 3,205 8,644 1,809 13,732 Feed Grains 113 3,142 1,948 1,249 6,452 Oilseeds 68 928 1,257 628 2,881 Other 68 244 402 550 1,264 Fallow 61 2,035 6,327 441 8,864 Total Cropland 383 9,553 18,579 4,677 33,192 Table 4.3 also summarizes the distribution of cropland i n western Canada. Saskatchewan has the largest land base with 56$ of total cropland, Alberta and Manitoba have 29$ and 14$ respectively, and B.C. contributes a further 1$, mainly in the Peace River region. Note that only 69% of the total cropland i s actively i n production, with the residual l e f t fallow. Of the cropland planted, 56% i s i n wheat, i l l u s t r a t i n g the relative importance of this crop. The remainder i s seeded to feed grains (27%), oilseeds (12$), and other minor crops (5$). Since 63$ of land planted to wheat i s i n Saskatchewan, that province i s l i k e l y to be the most affected by the introduction of medium quality wheat. Alberta, Manitoba, and B.C. account for 23$, 13$, and 0.5$, respectively, of the wheat acreage on the prairies. Saskatchewan has a higher proportion of i t s land i n wheat while Alberta produces more feed grains, reflecting the importance of the livestock industry i n the latter province. 98 4.3.2 Impaot of Medium Quality Wheat on Production The impact of medium quality wheat introduction on seeding patterns w i l l be analyzed i n this section. Effects at the provincial level w i l l be reviewed f i r s t , followed by an analysis at the prairie level. Acreage response and supply curves w i l l be estimated for medium quality wheat, and supply e l a s t i c i t i e s calculated. Finally, a brief overview of the results at the regional level w i l l be provided. 4.3.2.1 Production Effects at the Provincial Level Acreage changes for the four western provinces w i l l be presented i n this section. The impacts on seeding i n Alberta, Saskatchewan, and Manitoba are reported i n Table 4.4 and these results w i l l be discussed i n turn, concluding with a note on the effects i n the Peace region of Bri t i s h Columbia. Medium quality wheat f i r s t appears i n Alberta when the price i s $l45/tonne. At this price, 28$ of traditional wheat i s replaced by the high yielding variety, and as the relative price increases, the proportion of medium quality wheat expands u n t i l i t comprises 94$ of wheat acreage. The new variety gains mainly at the expense of hard wheat i n Alberta. However, when the relative price exceeds 0.88, two other effects appear. F i r s t l y , the acreages of feed grains and oilseeds decline s l i g h t l y , and secondly, the amount of fallow land increases by 2%. Recall that the quantity of a l l crops i s affected by the crop and fallow ratios selected. Although fallowing land involves a substantial cost i n terms of foregone production, the yields for subsequent crops are increased. If the yield of a crop grown on fallow land i s sufficiently high, the additional revenue may justify the additional cost. 99 Table 4.4: Change i n Area Planted to Each Crop at Various Price Levels of MQW - Alberta, Saskatchewan, Manitoba (Thousand Hectares) Price of MQW Change in Area Market Relative Hard Feed Price Price a Wheat MQW Grains Oilseeds Fallow ($/tonne) Alberta: 135 0.72 0 0 0 0 0 140 0.75 0 0 0 0 0 145 0.78 -893 893 0 0 0 150 0.80 -1454 1454 o 0 0 155 0.83 -2717 2717 0 0 0 160 0.86 -2717 2717 0 0 0 165 0.88 -3011 3004 -31 -2 41 170 0.91 -3006 3019 -44 -9 41 Saskatchewan: 135 0.72 -1242 1340 0 -98 0 140 0.75 -2496 2635 -41 -98 0 145 0.78 -3194 3333 -41 -98 0 150 0.80 -6217 6355 -41 -98 0 155 0.83 -7143 7281 -41 -98 0 160 0.86 -8000 8262 17 -279 0 165 0.88 -8000 8262 17 -279 0 170 0.91 -7997 8277 -1 -279 0 Manitoba: 135 0.72 0 0 0 0 0 140 0.75 -316 316 0 0 0 145 0.78 -505 505 0 0 0 150 0.80 -1555 1596 -65 24 0 155 0.83 -1555 1616 -99 32 7 160 0.86 -1761 1857 -157 43 19 165 0.88 -1761 1885 -212 71 19 170 0.91 -1761 1885 -212 71 . 19 a Price of MQW divided by price of high quality wheat 100 As noted earlier, Saskatchewan includes 56$ of a l l prairie cropland and 63$ of a l l traditional wheat production. At $135/tonne, medium quality-wheat replaces 14$ of hard wheat acreage and 8$ of oilseed crops. The acreage of the new variety then increases sharply with price, levelling off at 8.3 million hectares (97$ of wheat land) when the price i s $l60/tonne. Medium quality wheat gains are mostly at the expense of hard wheat i n Saskatchewan, although there i s generally a negative impact on a l l other crops throughout the price range. There i s a slight aberration i n the pattern when price i s between $160/tonne and $l65/tonne, probably related to the crop ratio selected. Turning to the results for Manitoba, the third largest wheat producer, medium quality wheat f i r s t appears profitable when price i s $l45/tonne (relative price 0.75), replacing 17$ of traditional wheat production. The new variety expands rapidly at the expense of hard wheat and feed grains as the relative price rises. When the price i s equal to $l60/tonne, 97$ of traditional wheat production has been replaced by medium quality wheat. Some hard wheat i s s t i l l required for provincial use, although this requirement could be met through imports from a surplus province by incurring a shipping cost. With a medium quality wheat price greater than $150/tonne, total wheat acreage i n Manitoba exceeds that of the base case, peaking at 1.9 million hectares (7$ increase). Concurrently, feed grain acreage gradually declines by 17$ while oilseed production actually expands by as much as 11$. The latter result may be surprising. However, bear in mind that medium quality wheat can only compete directly with hard wheat. The proportion of cropland i n feed grains and oilseeds i s determined completely by the cropping and fallow ratios selected. With price at $150/tonne, a new set i s 101 chosen such that both wheat and oilseed production are increased. Although the ratio constraints are restrictive and may appear to be a r t i f i c i a l , they are based on plantings over a five year period, and as noted earlier represent a "reduced form" approach i n attempting to endogenize rotation choices. Finally, medium quality wheat can also be grown i n the Peace River region of British Columbia. However, as B.C. has only 1$ of prairie cropland, and only 0.5$ of traditional wheat acreage, i t s possible contribution to medium quality wheat production w i l l be small. The results indicate that i t w i l l only be grown at relatively high prices (in excess of $l45/tonne). Providing this condition i s met, 50,000 hectares of hard wheat are replaced by medium quality wheat with the balance (24,000 hectares) retained for provincial requirements. No other changes were reported. 4.3.2.2 Production Effects at the Prairie Level The total effects on production for Western Canada are found by aggregating the results of the four provinces, and these are presented i n Table 4.5. Trends identified at the provincial level are s t i l l evident although the responses are now smoother as a result of aggregation. Medium quality wheat i s grown over the entire range of prices. As the price rises, i t increases from 10$ of wheat acreage at $130/tonne to 94$ at $170/tonne. The total amount of land allocated to wheat also expands, although slowly. In the extreme (relative price of 0.91), total wheat acreage increases by 3% over the base case. Although medium quality wheat gains primarily at the expense of hard wheat, overall i t has a s t r i c t l y negative effect on a l l other grains, and the strength of this effect generally increases with price. An exception i s oilseed production which recovers slightly when 102 price i s between $150/tonne and $155/tonne. In this interval, oilseed acreage actually starts to exceed the base level in Manitoba. However, this positive reaction i s overwhelmed by the strong negative effect in Saskatchewan, especially as the price of medium quality wheat continues to rise. Table 4.5: Change i n Area Planted to Each Crop at Various Price Levels of MQW - Western Canada (Thousand Hectares) Price of MQW Change i n Area Market Relative Hard Feed Price Price Wheat MQW Grains Oilseeds Fallow ($/tonne) 135 0.72 -1242 1340 0 -98 0 140 0.75 -2812 2950 -41 -98 0 145 0.78 -4593 4731 -41 -98 0 150 0.80 -9277 9455 -105 -74 0 155 0.83 -11465 11665 -139 -66 7 160 0.86 -12529 12887 -140 -235 19 165 0.88 -12822 13201 -227 -210 60 170 , 0.91 -12815 13231 -257 -217 60 a Price of MQW divided by price of high quality wheat. Finally, as the price exceeds $155/tonne, the amount of fallow land increases slightly. This change i s less than 1$ when compared with that of the base case. The increase i n revenue due to the higher price of medium quality wheat compensates for the costs incurred with fallowing additional land. The rate of adoption of medium quality wheat i s summarized i n Table 4.6. From these values, the acreage response curves can be plotted for each western province, and for the prairies i n aggregate (Figure 4.1). Note the 103 acreage response curve for B.C. i s not included due to the low level of production i n that province. The main results displayed in Figure 4.1 f a l l into two categories: (i) the relative amount of medium quality wheat planted i n each province, and ( i i ) the relative price at which the new variety f i r s t appears i n each province, mainly, 0.72, 0.75, 0.78, and 0.80 for Saskatchewan, Manitoba, Alberta, and B.C. respectively. Table 4.6: Total Area Planted to MQW at Various Price Levels by Province (Thousand Hectares) Price of MQW Area Planted to MQW Market Price Relative Price B.C. Alberta Saskatchewan Manitoba Western Canada Total ($/tonne) 135 140 145 150 155 160 165 170 0.72 0.75 0.78 0.80 0.83 0.86 0.88 0.91 0 0 0 50 50 50 50 50 893 1454 2717 2717 3004 3019 0 0 1340 2635 3333 6355 7281 8262 8262 8277 316 505 1596 1616 1857 1885 1885 0 1340 2950 4731 9455 11665 12887 13201 13231 a Price of MQW divided by price of high quality wheat 104 Figure 4.1: Acreage Response Curves for Medium Quality Wheat Price of MQW ($/tonne) Manitoba Alberta Saskatchewan Western Canada 130 \ , • 1 1 -r * 0 2 4 6 8 10 12 14 Area Planted to MQW (million hectares) 105 The distribution of medium quality wheat acreage corresponds closely with the relative distribution of wheat production in the base case (Section 4.3.1.), especially when the price i s above $155/tonne. Saskatchewan i s the largest producer, ranging from 6 3 $ of the medium quality acreages when price i s $170/tonne to 100? when price i s $135/tonne. The acreages i n Alberta and Manitoba are similar although Alberta's production i s usually greater. The acreage i s B.C. (approximately 0.5$ of the total) i s negligible compared to that of the other three provinces. The model solution reports the acreages planted to each crop by region, as well as the export levels for each crop. Based on these results, the regional production i n tonnes can be calculated, which i n turn can be aggregated to give the total production for western Canada. The total production and export levels of medium quality wheat for various price levels are noted i n Table 4.7, and plotted i n Figure 4.2. The difference between production and export levels (1.8 million tonnes) constitutes the quantity required for livestock feed i n Canada. As price rises, domestic requirements (represented by the horizontal distance between the two graphs in Figure 4.2) decline as a proportion of production. Comparing the supply curves for the new variety i n Figure 4.2 with the aggregate acreage response curve i n Figure 4.1, three segments can be. identified i n each: 1. a steep portion from $135/tonne to $l45/tonne during which medium quality wheat i s moderately competitive with hard wheat; 2. a fla t t e r portion from $l45/tonne to $155/tonne when the new variety i s very competitive; 3 . a steep portion from $155/tonne to $170/tonne when the remaining cropland available to medium quality wheat i s severely limited. 106 Table 4.7: Total Production and Exports of MQW at Various Price Levels (Thousand Tonnes) Price of MQW Market Relative Total Total Price Price a Production Exports ($/tonne) 135 0.72 3088 1314 140 0.75 6745 4971 145 0.78 11379 9606 150 0.80 21775 20001 155 0.83 26628 24855 160 0.86 29477 27704 165 0.88 30221 28448 170 0.91 30589 28816 a Price of MQW divided by price of high quality wheat Arc e l a s t i c i t i e s were calculated for each of the three segments of the acreage response and export supply curves, and the results are presented i n Table 4.8. Note, the third segment has been truncated at $l65/tonne to equate the lengths of the price intervals. Table 4.8: Acreage Response and Export Supply E l a s t i c i t i e s for MQW MQW Price Interval Acreage Export Market Relative Response Supply Price Price Elasticity Elasticity ($/tonne) 135 - 145 0.72-0.78 16.0 23.0 145 - 155 0.78-0.83 11.0 11.0 155 - 165 0.83-0.88 1.9 2.1 a Price of MQW divided by price of high quality wheat 107 Figure 4 .2 : Supply Curves for Medium Quality Wheat - Western Canada Price of MQW ($/tonne) 170 165 160 155 150 145 140 135 130 9 #• Amount exported Amount produced 10 15 20 25 30 35 Amount of MQW ( m i l l i o n tonnes) 108 The results indicate that acreage response and export supply are elastic throughout the price range, but particularly at the low end. For each elas t i c i t y , the large value for the f i r s t segment compared to the second may be surprising considering the relative slope of the corresponding curve. However, this emphasizes that i t i s the proportional, rather than the absolute, changes that matter i n calculating e l a s t i c i t y . When a small amount of medium quality wheat i s grown, proportional changes tend to be large resulting i n a large e l a s t i c i t y . With any new technology, the level of adoption i s i n i t i a l l y zero, and therefore, a highly elastic response could be expected when the technology i s f i r s t introduced. At the higher price levels, the export supply el a s t i c i t y i s similar to that of the acreage response el a s t i c i t y but at lower price levels they differ significantly. This difference i s due to the decline i n the proportion of production retained for domestic use as price increases. The results in Table 4.8 show that supply e l a s t i c i t i e s vary considerably with the price of medium quality wheat. Although the ultimate export price of medium quality wheat i s debatable, i t w i l l probably be i n the range $135/tonne to $155/tonne. This estimate i s based on the price of comparable Australian and U.S. wheats. Consequently, there i s a strong indication that the supply of the new variety w i l l be quite elastic. The foregoing analysis requires qualification. The e l a s t i c i t i e s reported in Table 4.8 are partial e l a s t i c i t i e s , i.e. they estimate the proportional change i n quantity supplied for a given proportional change i n the price of medium quality wheat, ceteris paribus. However, the quantity supplied w i l l depend not only on the price of medium quality wheat, but also on other variables such as the price of close substitutes i n production. By assumption, hard wheat i s the closest substitute for medium quality wheat, 109 and therefore, a more accurate index of supply response w i l l be provided by the relative price e l a s t i c i t y . 4.3.2.3 Production Effects at the Regional Level A knowledge of the regional impact of introducing medium quality wheat provides useful information regarding the distribution of benefits and future grain handling requirements. In the context of the current study, regional responses also help to explain the particular shape of the acreage response and supply curves (Figures 4.1 and 4.2). The pattern of adoption of the new variety w i l l vary with i t s relative price. When the price i s low, medium quality wheat i s only moderately competitive with traditional varieties. For most crop regions, the increased yields are insufficient to compensate for the significantly lower price. At $l40/tonne, the price of medium quality wheat i s 25% lower than the price of hard wheat, and the new variety w i l l only be grown i n those regions where the yield advantage i s particularly large. As the price of medium quality wheat increases, i t becomes more attractive to farmers in an increasing number of regions. The acreage response curve for western Canada (Figure 4.1) indicates that the relative price range from O.78 to 0.83 includes the breakeven point for many of the regions. When the relative price i s high (greater than 0 .86), only those areas particularly adapted to hard wheat w i l l be growing i t exclusively. The regional responses to three different prices ($l40/tonne, $150/tonne, and $l60/tonne) are given i n Table 4.9. Note, the prices chosen correspond to the mid-point values of the three segments used in calculating the e l a s t i c i t i e s i n the last section. To better visualize the change in the adoption pattern as price increases, three maps of the prairie crop regions Table 4.9: Total Area Planted to MQW at Various Price Levels by Region (Thousand Hectares) Area planted to MQW  Crop Dominant Region Location Soil Zone $140/Tonne $150/Tonne $160/Tonne (thousand hectares) B.C. 1 Peace (Dawson Creek) Gray Luvisol 0 50 50 Alberta 1 south-east (Medicine Hat) Brown 0 0 0 2 south (Vulcan) Dark Brown 0 0 1002 3 south-west (Olds) Gray Luvisol, Black 0 0 261 4 east (Viking) Dark Brown, Black 0 801 801 5 central (Red Deer) Gray Luvisol, Black 0 93 93 6 north (Barrhead) Gray Luvisol 0 95 95 7 Peace (Falher) Gray Luvisol 0 466 466 Sask. 1 south-east (Grenfell) Dark Brown, Black 0 747 747 2 south-east central (Weyburn) Brown, Dark Brown 0 0 926 3 south-west central (Moose Jaw) Brown 0 1588 1623 4 south-west (Fox Valley) Brown 0 0 0 5 east (Foam Lake) Black 1380 1380 1380 6 central (Saskatoon) Dark Brown 1255 1255 1255 7 west (Rosetown) Brown, Dark Brown 0 0 857 8 north-east (Melfort) Black, Gray Luvisol 0 698 685 9 north-west (Prince Albert) Black, Gray Luvisol 0 688 789 Manitoba 1 south-west (Brandon) Black 0 597 633 2 north-west (Grandview) Black, Gray, Rich Lime 316 316 316 3 central (Swan River) Black 0 0 206 4 south (Elm Creek) Black 0 348 348 5 south-east (Winnipeg) Gray, Rich Lime, Black 0 189 210 6 interlake (Arborg) Rich Lime 0 145 145 111 have been prepared based on the data i n Table 4.9. Figure 4.3 corresponds to a price of $l40/tonne for medium quality wheat, Figure 4.4. to $150/tonne, and Figure 4.5 to $l60/tonne. Recall from Chapter one that the particular crops grown i n a location depend largely on s o i l and climate. To enable the rate of adoption of the new wheat variety to be related to one particular s o i l characteristic (i.e. colour), a map of prairie s o i l zones i s included (Figure 4.6). In addition, the correspondence between s o i l zone and crop region i s given i n Table 4.9. By comparing Figure 4.6 with the three crop region maps, one can detect a trend; medium quality wheat tends to be adopted f i r s t by regions i n the black, dark brown, and gray s o i l zones, followed by areas i n the brown zone. The relative yield advantage i s greater i n the black s o i l zone than i n the brown zone. However, the correspondence between s o i l zone and rate of adoption i s not perfect. Other important s o i l characteristics such as texture (which affects water holding capacity) are ignored i n this broad categorization. For instance, thin black soils are generally poor for wheat, whereas the thick black soils such as those occurring in the Melfort-Tisdale (Saskatchewan region 8) and Red Deer-Lacombe (Alberta region 5) d i s t r i c t s are relatively good. The adoption pattern may be examined more closely. When the price of medium quality wheat i s $l40/tonne (Figure 4.3), the new variety i s grown in a band which includes the eastern part of central Saskatchewan (regions 5 and 6), and the western part of central Manitoba (region 2). These regions consist largely of black and dark brown so i l s . 112 Figure 4.3: Prairie Regions Adopting MQW 'at $140/Tonne OUTLIMC UA * o r i w f p n i i n i c p n o v . N C C S Figure 4.4: Prairie Regions Adopting MQW at $150/Tonne Figure 4.5: Prairie Regions Adopting MQW at $160/Tonne Figure 4.6: Generalized S o i l Map of the Prairie Provinces (Horner et a l . , p. 5) 114 When the price rises to $1507tonne (Figure 4.4), medium quality wheat i s adopted by every region except regions 1, 2, and 3 in southern Alberta, regions 2, 4, and 7 in south-western and south-central Saskatchewan, and region 3, south of Portage La Prairie. Note, the heavy clay area around Regina-Weyburn (Saskatchewan region 2), and the clay-loam soils around Rosetown-Kindersley (Saskatchewan region 7) are two of the best wheat growing areas i n Canada. Finally, at $l60/tonne (Figure 4.5), the only regions not growing medium quality wheat, south-eastern Alberta (region 1) and south-western Saskatchewan (region 4), are part of the brown s o i l zone. The pattern of adoption described i n the current study corresponds to the conclusions of Ulrich and Furtan. They analyzed the mean grading pattern of CWRS (hard) wheat i n each of the crop regions. Their " i n i t i a l results showed that the economic incentive to grow HY320, expressed as the percent increase i n farmgate wheat sales, depended most upon the historic grading pattern an area experienced. The historic grading pattern was, in turn, associated with the s o i l zone the area was i n . Generally, the brown-soil zone (high-grading) crop d i s t r i c t s had less incentive to adopt HY320 than the black and gray (low-grading) crop d i s t r i c t s . " (Dlrich and Furtan, p. 55). 4 .3 .3 . Impact of Medium Quality Wheat on Export Levels Estimates of export volumes for different relative prices of medium quality wheat are useful for a number of reasons. They provide an estimate of: (i) expected strain on the existing grain delivery system; ( i i ) expected export earnings; and 115 ( i i i ) expected market share. An estimate of market share i n turn would allow analysts to evaluate the risk of retaliation by competing nations. This section examines the impact of introducing the new variety on wheat exports, and total grain exports. Table 4.10 provides a summary of wheat exports by class. The results indicate that both total volume and class composition are affected. The change i n total wheat exports s t r i c t l y increases with price. The increment varies from 5$ at $135/tonne to 37$ at $170/tonne. Over this price range, the proportion of medium quality wheat increases from 6% to 98$, whereas hard wheat exports decline steadily. They are down by 1$ of the base level at $135/tonne, and by 98$ at $170/tonne. Table 4.10: Summary of Wheat Exports by Class at Various Price Levels of MQW Price of MQW Total Wheat Exports Change i n Total Wheat Exports from Base Case Composition by Class Change i n Hard Wheat Exports from Base Case Market Relative Price Price Hard Wheat MQW ($/tonne) (thou, tonnes) ($) ($) ($) ($) 135 0.72 22472 5.3 94.2 5.8 -0.9 140 0.75 23529 10.2 78.9 21.1 -13.1 145 0.78 24653 15.5 61.0 39.0 -29.5 150 0.80 27006 26.5 25.9 74.1 -67.2 155 0.83 27949 30.9 11.1 88.9 -85.5 160 0.86 28701 34.5 3.5 96.5 -95.3 165 0.88 28887 35.3 1.5 98.5 -97-9 170 0.91 29255 37.1 1.5 98.5 -97.9 Price of MQW divided by price of high quality wheat Total wheat exports i n base case = 21,345,000 tonnes 116 Table 4.10 allows a useful comparison with Henning's results. He developed a price endogenous econometric model of the world wheat market to analyze price and revenue effects resulting from a change i n the class composition of Canadian wheat exports. Henning varied the proportion of hard wheat replaced by an appropriate amount of medium quality wheat, and then solved for price, import levels, and non-Canadian export levels for each of the four classes of wheat i n the model. Of the five replacement levels examined (10$, 20$, 30$, 40$ and 50$), Henning concluded that "a 40$ re-orientation would be a reasonable upper lim i t i n the consideration of changes i n export composition" (Henning, p. 332). Loyns and Carter i n their study assumed that approximately 40$ of hard spring wheat acreage would be shifted into the new variety (Loyns and Carter, p. 9). With this level of substitution, Henning's model predicted a relative price for medium quality wheat i n 1984 of 0.78 (Henning, p. 320). These values correspond closely with the results of the current study where prices are exogenous but export levels are endogenized. Referring to Table 4.10, the mean decline i n exports of hard wheat i s 40$ over the relative price range 0.75 to 0.80. Dsing this price range, total wheat exports could rise by 10$ to 26$, with medium quality wheat comprising from 21$ to 74$ of these exports. The introduction of the new variety w i l l also affect the export levels of other crops. Table 4.11 summarizes the increase i n total grain exports over the price range analyzed. As indicated, the change i n aggregate exports i s s t r i c t l y increasing with price. Using the price interval suggested by Henning's results ($l40/tonne to $150/tonne), export tonnages could rise by 7$ to 19$, possibly straining the current grain handling system, but presumably increasing export earnings (Section 4.3.4.2). Comparing Tables 4.10 and 4.11, the percentage increase in wheat 117 exports consistently exceeds the increase i n total exports of grain. The clear implication i s that the introduction of medium quality wheat w i l l have a small but adverse effect on the other major crops. This was also recognized i n the analysis of production effects at the prairie level (Section 4.3.2.2). Table 4.11: Summary of Increases i n Exports from Prairie Provinces of Major Grains with Introduction of MQW Price of MQW Market Relative Price Price Increase i n Grain Exports from Base Case ($/tonne) (thou, tonnes) ($) 135 0.72 1038 3.6 140 0.75 2020 7.1 145 0.78 3144 11.0 150 0.80 5358 18.7 155 0.83 6233 21.8 160 0.86 6843 23.9 165 0.88 6855 24.0 170 0.91 7045 24.6 Price of MQW divided by price of high quality wheat Total exports i n base case = 28,592,000 tonnes Henning reported that between 1979 and 1981, world trade i n class 2 wheat averaged approximately 36.3 million tonnes (Henning, p. 60). Using this as a reference, the export levels i n Table 4.10 could constitute a sizeable market share, even at comparatively low relative prices. For example, over the price range $l40/tonne to $150/tonne, medium quality wheat exports could vary between 5 and 20 million tonnes. This corresponds to a market share between 14$ and 55$ which would infringe upon the wheat sales 118 of major competitors. The U.S. share of total wheat exports i s approximately 41$ (Veeman and Veeman, p. 27). Approximately half the total U.S. wheat crop i s medium quality wheat (Ulrich and Furtan, p.40). Thus, price retaliation could be a very real danger. In any case, the assumption that demand i s i n f i n i t e l y elastic could be seriously questioned. 4.3.4 Impact of Medium Quality Wheat on Income Although the mean yield for medium quality wheat i s appreciably higher than for traditional varieties, the price received w i l l be lower. The commercial success of the new variety w i l l depend very much on i t s ultimate price as i t s adoption w i l l only be justi f i e d i f net income i s increased. Income effects can be measured i n different ways. This section w i l l evaluate the impact of introducing the new variety on three income measures: 1. change i n net farm income 2. change i n export earnings 3. change i n sectoral earnings. 4.3.4.1 Change i n Net Farm Income Change i n farmgate sales can be calculated directly by comparing results for a given price range with those of the base case. But, as Ulrich and Furtan point out, change i n farmgate sales i s equivalent to change i n net farm income since variable costs of production have been deducted, and fixed costs w i l l have been paid i n any case (Ulrich and Furtan, p. 64). The changes in net farm income for prairie grain producers are summarized i n Table 4.12. Two results are readily apparent. First, change in income i s positive (or more generally, non-negative) for a l l values of price. Generally, a farmer cannot suffer i n absolute terms from having a 119 wider range of planting options. Second, additional income increases with price, varying from $9 million to $715 million over the entire range. For the restricted price range $140/tonne to $150/tonne, the additional income varies from $34 million (2$ increase) to $155 million (9% increase). Thus, the results indicate that overall, medium quality wheat i s relatively profitable. Table 4.12: Changes in Net Farm Income of Prairie Grain Producers from Introducing MQW Distribution of Price of MQW Benefits by Province Market Price Relative Price Change i n Net Farm Income . of Prairie Grain Producers B.C. Alta. Sask. Man. ($/tonne) ($ million) % (%) 135 0.72 9 0.5 0 0 100 0 140 0.75 34 2 0 0 97 3 145 0.78 74 4 0 8 86 6 150 < 0.80 155 9 0.5 15 74 11 155 0.83 272 16 0.6 17 69 13 160 0.86 412 24 0.6 21 65 13 165 0.88 563 33 0.6 22 63 14 170 0.91 715 41 0.6 23 62 14 a Price of MQW divided by price of high quality wheat Total farmgate sales i n base case i s $4,058 million. Realized net income from farming operatioNs on prairies i n 1982 was $1,728,000,000 (Ulrich and Furtan, 1984, p. 65; Statistics Canada # 21-202). Clearly, any estimate of net income effect i s sensitive to assumptions regarding price. Loyns and Carter calculated that net producer gains would be between 5% and 17$ (Loyns and Carter, p. 17). According to the current study, these values correspond to a price range of $l45/tonne to $155/tonne. Dlrich and Furtan, on the other hand, estimated the gains to be 120 between 14$ and 24$ (Ulrich and Furtan, p. 65). For benefits of this magnitude, the price i n the current analysis would be between $150/tonne and $l60/tonne. It would be desirable to compare the prices used i n the various studies. Unfortunately, this i s complicated due to the different approaches used. Loyns and Carter employ a simple price endogenous model i n which the resulting price i s a single blended one for a l l categories and grades of wheat. In Ulrich and Furtan 1s accounting type model, the price of wheat i s exogenous as i t i s i n the current study. However, they differentiate high quality (class 1) and medium quality (class 2) wheat into four grades and three grades respectively, each with a unique price. By contrast, the current study uses a blended price for each of the top two classes of wheat. Table 4.12 indicates how the benefits of introducing medium quality wheat may be distributed across provinces. The distribution correlates well with the acreage response results over the range of prices (Figure 4.1). A province can only directly benefit from the new variety i f the province grows the wheat. The distribution of gains within a province would be determined by the adoption rate of each region (Section 4.3.2.3). When the price of medium quality wheat i s $l45/tonne, 86$ of the resulting benefits go to Saskatchewan, with Alberta and Manitoba sharing the residual relatively equally. As the price rises, Saskatchewan's share gradually declines, although i t continues to enjoy the largest portion. By contrast, the proportions accruing to Alberta, and Manitoba increase, although at a faster rate for Alberta, resulting i n a broadening of the spread between the two. Once B.C. starts producing, i t s share remains constant, but small (0.6$ of total benefits). Ultimately, an equilibrium i s approached i n which Saskatchewan gains by 62$, Alberta by 23$. Manitoba by 14$, and B.C. by 0.5$. This pattern closely matches the distribution of wheat production obtained i n the base case (Section 4.3.1). 121 4.3.4.2 Change i n Export Earnings The computation of export earnings i s based on the export level and price of each major grain. As discussed earlier, the introduction of medium quality wheat has an overall negative effect on the production of a l l other crops, especially high quality wheat (Table 4.5). Nevertheless, total wheat production increases. Since domestic requirements are essentially fixed i n the model, apart from some substitution po s s i b i l i t i e s i n livestock feed, most of the change i n production i s reflected i n the export data. The impact of introducing the new variety on the level of export earnings i s summarized i n Table 4.13. As would be expected, earnings s t r i c t l y increase with price ranging from a gain of $109 million (an increase of 2% over the base case) to a gain of $805 million (increase of 15$). Table 4.13: Change i n Export Earnings of A l l Grains from Introducing MQW Price of MQW Market Price Relative Price Change i n Export Earnings from Base Caseb ($/tonne) ($ million) (%) 135 0.72 109.2 2.1 140 0.75 132.6 2.5 145 0.78 174.2 3.3 150 0.80 266.7 5.0 155 0.83 381.7 7.2 160 0.86 514.4 9.7 165 0.88 653.3 12.3 170 0.91 804.7 15.1 Price of MQW divided by price of high quality wheat Export earnings i n base case = $5,313.5 million 122 Comparing Tables 4.12 and 4.13, the gain i n export earnings i s approximately $100 million more than the gain i n net farm income. This amount represents the additional costs of production and shipping, including the benefits accruing to the transportation sector. Since the earnings increase more rapidly than these costs over the price range of medium quality wheat, the proportion of benefits going to the producer must also rise. The analysis assumes that the increased volumes of grain could be shipped to port at the subsidized (Crow) rate. However, under the current system of payment, as outlined i n the Western Grain Transportation Act, farmers must pay the compensatory, or f u l l , rate for a l l volumes over 31.5 million tonnes per year (Lerohl, et a l . ) . If, in fact, farmers were required to pay the higher rate on the additional shipments associated with the higher yielding wheat, the reduced benefit to the producers would be partially offset by an increased benefit to the transportation sector. Note, however, there would also be a production effect as crops with a higher value to weight ratio would be favoured. The results i n Table 4.13 correspond with Henning's estimates of additional export earnings. As described i n Section 4.3.3, he analyzed the price and revenue effects resulting from a change i n the class composition of wheat exports by replacing various percentages (10$, 20$, 30$, 40$, 50$) of hard (class 1) wheat by the appropriate amount of medium quality (class 2) wheat. Henning estimated that for 1984 the increase i n export earnings would range from $100 million to $260 million over the five regimes examined (Henning, p. 332). For a re-orientation of 40$, the upper l i m i t suggested by Henning, export earnings would have increased by $250 million, or by 6$ over his base case level. In the current study, when the relative price i s 123 0.80 ($150/tonne), additional export earnings are $266 million, an increase of 5% over the base case level. Thus, the two studies agree closely in this area. Note, for the 40$ replacement regime, Henning estimated that the corresponding decrease i n revenues to the D.S. could be 6.1$ (Henning, p. 360). Considering that U.S. wheat shipments are twice the volume of Canadian exports, the threat of retaliation could be very real. (See also Section 4.3-3 for a discussion of market share.) The only commodity considered i n Henning's study was wheat although the crop was differentiated into four categories. His results demonstrated that the introduction of medium quality wheat w i l l have a positive effect on the price of class 1 (hard) wheat, but a negative impact on revenue earned by class 3 (soft) and class 4 (Durum) wheats. In the current study, wheat i s limited to two classes (hard and medium hard) although a broad range of crops, as well as major livestock categories, are included. The results indicate that the introduction of the new variety w i l l have an overall negative effect on the production and export levels of non-wheat crops (Table 4.5). 4.3.4.3 Change i n Agricultural Sector Earnings The value of the model's objective function i s equal to the net earnings of the agricultural sector. It i s the revenue earned by the crop and livestock production act i v i t i e s minus production and transportation costs. Consequently, the term "agricultural sector" i s used here to refer to a l l those production a c t i v i t i e s included i n the model. For a given price of medium quality wheat, the increase i n net sector earnings from the base case level i s simply the difference between the objective function values i n the two solutions. These results are tabulated i n Table 4.14. 124 Table 4.14: Change in Net Earnings of the Agricultural Sector from Introducing MQW Price Market Price of MQW Relative Price a Change i n Net Sector from Base Case bEarnings Change i n Net Sector Earnings Minus Change i n Net Income of Prairie Grain Producers ($/tonne) ($ million) (%) ($ million) 135 0.72 100.6 1.0 92 140 0.75 116.5 1.1 83 145 0.78 147.6 1.5 74 150 0.80 220.8 2.2 66 155 0.83 328.8 3.2 54 160 0.86 459.1 4.5 42 165 0.88 601.1 5.9 32 170 0.91 744.9 7.3 23 Price of MQW divided by price of high quality wheat Net Sector Earnings i n Base case = $10,201.7 million Comparing Table 4.14 with Table 4.13, the increase i n export earnings of the cropping a c t i v i t i e s consistently exceeds the additional net sector earnings, albeit by a relatively small margin. Although "export earnings" as defined do not include revenue from domestic sales or from the foreign sales of livestock products, neither have they been adjusted for production and transportation costs. The implication i s that the revenue from sources other than grain exports i s insufficient to offset a l l production and transportation costs. Thi3 reiterates the importance of grain to the agricultural sector. As the price of medium quality wheat rises, the acreage planted increases resulting i n additional production and transportation costs associated with the higher yields. Consequently, the gap between increased sector and export earnings widens. 125 In Table 4.14, a comparison between changes i n net sector earnings and changes in net farm income i s presented. For each price level, the additional sector earnings are greater, and the difference represents the net benefits accruing to the livestock sector, and to crop production i n eastern Canada. As shown, these production a c t i v i t i e s experience substantial gains when the price of medium quality wheat i s low ($92 million at $135/tonne). As the price increases, this net benefit diminishes. Recall that the main domestic use of the new wheat variety w i l l be as livestock feed. Consequently, the lower the price, the more attractive i t w i l l be as a substitute for traditional feed grains. Finally, i n percentage terms, the increase i n net sector-wide earnings i s considerably lower than for net farm income or for export earnings, but this i s only because the base i s considerably higher. For example, export earnings (crops only) i n the base case are $5 b i l l i o n as compared to $10 b i l l i o n i n net revenue for the entire agricultural sector. 4.3.5 Relaxing the Constraint on Medium Quality Wheat In the analyses so far, medium quality wheat was restricted to competing with hard wheat for land allocated to wheat production. This constraint limits the level of adoption of the new variety, and therefore, the resulting benefits. It i s of interest to examine the response of the model when this restriction i s relaxed. Consequently, the price scenarios were repeated allowing medium quality wheat, hard wheat, and barley to compete directly for land jointly available for wheat and barley production. As before, barley i s used to represent a l l feed grains, and the two terms w i l l be used synonymously in this section. For the original set of experiments, the results were presented in 126 terms of "change from the base case" where "the base case" was that solution resulting from fixing the level of medium quality wheat at zero, i.e. the new variety was not a planting option. For the current set of tests, the old base case i s no longer suitable. To explain, consider an increase i n the level of production of a grain. The resulting benefit would have two components. F i r s t l y , the benefit due to relaxing the constraint on hard wheat and barley, and secondly, the benefit due to introducing medium quality wheat. The second effect i s the one of interest. To remove the f i r s t component, a new base run was made in which hard wheat and barley could compete for the total cropland available to these two crops. One could expect some adjustment i n relative production levels due to differences i n pr o f i t a b i l i t y based on the exogenous price, cost, and yield data. Ideally, no adjustment would occur i n the solution. However, i t i s recognized that not a l l constraining factors have been modelled. The aggregated results of the original and revised base case solutions are compared i n Table 4.15. As indicated, wheat land increases by 1.5 million hectares (+11$), at the expense of barley (-20$), oilseeds (-8$), and some fallow land. The reduction i n the level of oilseed and fallow acreages i s the result of a change i n the optimal cropping and fallow ratios selected. Although the discrepancies may be acceptable at the aggregate level, the results are sufficiently different for some regions (not shown) to question the validity of some of the input data used, and the f a i r l y aggregate level of constraints specified at the regional level. Recall that the model was validated using the original base case as outlined i n Section 4.1. An alternative approach would have been to extend the substitution possibilities of medium quality wheat by allowing i t to compete with hard 127 wheat on land available for wheat production, and with barley on land available for barley production. In other words, only allow hard wheat and barley to compete indirectly with each other through the optimal choice of cropping mix. This approach would have an important advantage i n that the original validated base case would s t i l l be applicable thus allowing the two sets of results to be directly compared. Although the necessary modifications would not be extensive, time did not permit this analysis. Table 4.15: Comparison of Base Case Results under Two Substitution Options for MQW Area Planted to Western Each Crop i n Canada Grain Base Case #1B Basg Case Change i n Area from Base Case #1 to Base Case #2 (thou, ha.) (thou, ha.) (thou, ha.) ($) Hard Wheat Feed Grains Oilseeds Other Fallow 13732 6452 2881 1264 8864 15287 5157 2653 1264 8832 +1555 +11.0 -1295 -20.0 -228 -8.0 0 -32 -0.4 Total Area 33192 33192 a Base Case #1: Hard wheat and medium quality wheat oompete directly. Base Case #2: Hard wheat, medium quality wheat, and feed grains compete directly. Production effects at the prairie level are reported i n Table 4.16. Note, the corresponding results for the more restrictive case are displayed i n Table 4.5. As shown, medium quality wheat i s grown at a l l prices. At the lowest price, 10$ of total wheat acreage consists of the new variety, but this proportion increases steadily with price so that by $170/tonne, 98$ of total wheat acreage i s comprised of the high yielding wheat. 128 Table 4.16: Change i n Area Planted to Each Crop i n Western Canada From Base #2 at Various Price Levels of MQW - Constraint on MQW Relaxed (Thousand Hectares) Price of MQW Change i n Area Market Relative Feed Price Price Hard Wheat MQW Grains Oilseeds Fallow ($/tonne) 135 0.72 -977 1619 -544 -98 0 140 0.75 -2637 2843 -75 -130 0 145 0.78 -6302 6701 -269 -130 0 150 0.80 -11329 11964 -426 -210 0 155 0.83 -13543 14377 -459 -375 0 160 0.86 -14306 15347 -510 -531 0 165 0.88 -14988 16029 -510 -531 0 170 0.91 -14988 16042 -510 -544 0 a Price of MQW divided by price of high quality wheat The acreage of medium quality wheat expands at the expense of a l l other grains, although the negative effect i s greatest for hard wheat. At $145/tonne, traditional wheat production i s down by 41$ which i s essentially the upper bound on the replacement level suggested by Henning (Henning, P. 332). Ultimately, the acreage of traditional wheat varieties i s down by 98$, and only 299 thousand hectares are seeded. Despite this decline, total wheat acreage increases steadily with price. The percentage of total cropland seeded to wheat expands from the base level of 46$ to 49$ when the price i s $170/tonne, a maximum increase of 7$ in wheat production. Other crops grown are also adversely affected by the introduction of medium quality wheat. Feed grain production i n i t i a l l y drops sharply as the new variety becomes a very attractive alternative for livestock feed. When the price of medium quality wheat rises to $l40/tonne, i t i s less attractive relative to barley, and consequently, feed grain production recovers considerably. However, as the price continues to increase, feed grain 129 production gradually declines. Eventually, feed grain acreage i s 10? lower than the base level. When compared to the original set of results (Table 4.5), feed grain production i s considerably more responsive to the price of medium quality wheat since the two types of grain now compete directly on the supply side. As before, the new variety can only compete indirectly with oilseeds via the optimal choice of cropping and fallow ratios. Nevertheless, oilseed production f a l l s considerably over most of the price range, levelling off at a 20$ reduction i n acreage. The amount of fallow land, on the other hand, i s unaffected by the introduction of medium quality wheat. 4.4 Modifying Assumptions and Limitations Section 4.3 presented the impacts of the introduction of medium quality wheat organized under production, export, and income effects. These estimates must be interpreted i n the context of the model used to derive them. A l l models are an abstraction of reality, and i n making simplifying assumptions, details are often suppressed which could affect these estimates. Six concerns are identified below. F i r s t l y , the prices which are set exogenously should be interpreted as expected prices that w i l l prevail, and based on these, a supply curve i s estimated. Given this supply curve, the equilibrium price and quantity would be determined by world demand for Canadian medium quality wheat, and the characteristics of this demand would be conditional on the potential retaliatory actions of major competitors. No indication i s given of the adjustment process that operates i n arriving at an equilibrium state. In particular, the short run Canadian response to i n i t i a l prices could correspond to a potential market share which would be unacceptable to other 130 exporting nations such as the United States. A second point concerns the effect of medium quality wheat production on the prices of other grain. This i s ignored but i f prices had been allowed to adjust, the reported changes would have been less dramatic. Henning noted that by replacing up to 50% of Canadian hard wheat exports by medium quality wheat, the price of hard wheat could rise by as much as 17.6$ (Henning, p. 320). Thus, farmers growing only hard wheat could s t i l l benefit from the new variety through i t s impact on price. Henning*s findings also indicate that the current results should be interpreted i n terms of the relative price (which this study did) rather than the absolute price of medium quality wheat. Third, the characterization of agronomic and biological constraints are highly simplified i n this study. Essentially, they are modelled by sets of cropping options which specify valid combinations of crops, and by fallow ratios which ensure that some land i s fallowed. The cropping ratios are based on observations i n each region over a five year period. These aggregate results are determined by the seeding decisions of individual farmers who did not have medium quality wheat as a planting option. Presumably i f the new variety had been available, the observed proportions of wheat, feed grains, and oilseeds could have been different. I n i t i a l l y , medium quality wheat was modelled to directly substitute for hard wheat only which assumes the two varieties are agronomically compatible i n terms of timing of operations, and rotational constraints. Recognizing that this approach may be too restrictive, this constraint was later relaxed by allowing direct competition of wheat and barley. The fourth consideration deals with the role of risk i n the analysis. The results of the model represent average values based on 1984 blended 131 prices and mean yields (assuming an average grade) over a five year period. However, yields, grades, and prices do fluctuate over time and consequently, risk considerations are important. A farmer associates a degree of risk with each crop based on historical observations, possibly adjusted subjectively. Compared with traditional varieties, HY320 may involve greater risk due to i t s longer maturation period, poorer resistance to disease, and greater susceptibility to sprouting (Table 3.4). Although a farmer's evaluation of risk for traditional crops may be reflected i n the cropping ratios, the greater risk associated with HI320 would not be captured. Consequently, assuming farmers are risk averse, i f medium quality wheat involves greater risk the acreage response levels of the new variety w i l l have been overestimated. Note, however, the degree of risk associated with medium quality wheat w i l l diminish with time as improved varieties are developed and licensed, and better management practices evolve. Fifth, i t i s assumed that the increased volumes of grain associated with the higher yielding wheat would be shipped at the subsidized (Crow) rate. However, under the current system of payment, as outlined i n the Western Grain Transportation Act, farmers must pay the compensatory, or f u l l , rate for a l l volumes over the record level achieved i n 1981/82. If this policy continues, crops with a higher value to weight ratio would be favoured, and consequently, the production level of medium quality wheat w i l l have been overestimated. Finally, the study focuses on the benefit of medium quality wheat to prairie grain farmers, largely ignoring the effects on other sectors, as well as the multiplier effects of increased earnings. Assuming the price of hard wheat increases, producers of high quality wheat i n eastern Canada would enjoy additional gains. Livestock producers in both eastern and 132 western Canada would benefit from the increased supply of feed quality wheat. Finally, the transportation sector would benefit from the increased shipments, especially i f the compensatory rate i s charged on volumes i n excess of the 1981/82 levels. Consequently, the economy-wide benefits of introducing medium quality wheat w i l l exceed the sector-based results focused upon i n this study. This section has focused on certain modifying assumptions that must be borne i n mind i f the results presented in Section 4.3 are to be examined. These concerns are important i n that an ever changing economic environment forces one to be cautious i n an analysis dealing with the introduction of a new technology where histor i c a l information i s limited. 133 CHAPTER 5 Summary and Conclusions In this chapter, the study i s summarized, and conclusions are drawn. The f i r s t section provides a motivation for the work and defines the objectives. The model of the Canadian agricultural sector developed i n Chapters 2 and 3 i s reviewed, i t s application to the current problem discussed, and the results are then presented. Section 2 draws certain policy implications from the results, and Section 3 provides suggestions for further research. 5.1 Summary of the Study Historically, Canada has concentrated on producing hard wheat. Through i t s licensing and inspection system this country has established a reputation as a reliable supplier of a high quality product, and consequently i s able to command a premium price for i t . However, over the past twenty years, growth i n world demand has been mainly centered on the softer varieties, contributing to a gradual erosion of Canada's market share. On the production side, i t appears that substantial increases i n yields w i l l only be achieved at the expense of quality. These two facts have resulted i n pressure on researchers to develop higher yielding medium quality wheats, and subsequently on the Canadian Grain Commission to license these new varieties. To be incorporated i n the inspection and marketing system, any new wheat must be visually distinguishable from those currently licensed. One medium quality wheat cultivar, HY320, has unique visual characteristics and 134 i t was licensed for production i n January, 1985. The objective of this study i s to examine the impact of this new variety on production opportunities and incomes for prairie farmers and to estimate the share of grain exports that i t may command i n the future. With the licensing of medium quality wheat, farmers must now decide whether increased yields w i l l more than compensate for the lower market price. Their seeding decisions w i l l determine aggregate Canadian output and export levels. In modelling this supply response sufficient detail must be included to adequately reflect the decision process. In studies concerning the impact of new technology there i s usually l i t t l e historical data of assistance to the model builder. Under these circumstances the mathematical programming methodology i s useful. A sectoral model i s developed i n which the country i s divided into regions for which aggregate ac t i v i t i e s and constraints are defined. The model i s sector-wide i n the sense that i t describes domestic supply and use of major crop and livestock commodities i n Canada. The problem i s to determine the level of agricultural production which maximizes net returns to the agricultural sector subject to constraints facing the sector. It i s a single period model but may be solved recursively to examine changes over any planning period of interest. The model contains approximately 1180 act i v i t i e s and 725 constraints. The a c t i v i t i e s can be divided into three major groups: production, shipping, and marketing a c t i v i t i e s . The production a c t i v i t i e s are further subdivided into crop and livestock blocks. A crop production block, or set of crop producing a c t i v i t i e s , i s defined for each of 29 regions, 22 of which are within the prairie provinces. Livestock blocks are defined at the provincial level for 7 different regions. The principal categories of crops 135 include grain, forage crops, pasture and minor crops. The grain category consists of wheat, barley, flax, canola, soybeans, and corn. The four livestock types Included are: beef, dairy, hogs, and poultry. Transport ac t i v i t i e s specify the shipment of grains, beef, and pork from production regions to domestic demand centres and to export points. Thus, interprovincial and international trade are modelled. The domestic demand block deals with the sale of crop and livestock commodities within Canada, and once domestic requirements are satisfied, the residual i s sold on the world market. The constraint rows i n the model deal mainly with land a v a i l a b i l i t y , and with opening and closing beef herd sizes. Commodity balance rows are specified for each major item and ensure that use cannot exceed supply. Accounting rows are principally used to record cash costs, and ratio rows have been specified to allow for alternative crop rotations, and the fallowing of land. Bounds or right hand sides specify upper levels on certain constraints, and are also used to specify activity levels for exogeneous variables. The exogeneous variables include: the size of the dairy, hog, and poultry a c t i v i t i e s ; the opening and closing stocks of beef animal classes; grain inventories; the level of certain import and export activites; and domestic demand levels. As the number of non-zero coefficients i n the model i s i n the order of 4900, the data requirements of the model are extensive. The values required include: land acreages, crops grown, and cost and yield data for crop production blocks; herd and flock sizes, diet, cost and yield data, and biological parameters for the livestock production block; commodities shipped, routes, and unit costs for the transportation block; and, f i n a l l y , 136 commodities sold, prices, and domestic levels for the marketing block. Many different data sources are referenced. Most .represent published documents from Statistics Canada, Agriculture Canada, and provincial ministries of agriculture. In addition, a deliberate attempt i s made to f u l l y u t i l i z e the National Farm Survey results for production data. In general, 1984 was taken as the base year for coefficients i n the model, but when i t was more desirable to use data pertaining to five year averages, the interval 1980 to 1984 was used. In special cases, the values applied to longer periods. To f a c i l i t a t e the manipulation of input and results, computer software was developed. A matrix generator program was written to format the input data correctly; a report writer was used to display the results i n convenient tables; and a tabulation program was written to aggregate the regional results to the provincial, and western Canada level. Once the model was developed and validated, i t was used to forecast the consequences of introducing HY320 (the variety of medium quality wheat recently licensed) into production. The rate of adoption of the new variety depends mainly on relative costs, yields, and prices. Data on comparative costs and yields of HY320 were assembled for each of the 23 prairie crop regions. The data indicate an average yield advantage of between 25 and 30$. Prices of traditional crops were set at their 1984-85 level based on Thunder Bay. As l i t t l e medium quality wheat has been sold by the Canadian Wheat Board, there i s considerable uncertainty concerning i t s ultimate price. Consequently, the analysis was performed at eight specific prices between $135/tonne (the lower range for American and Australian medium quality wheat traded on the world market) and $170/tonne (the f i n a l realized price for HY320 in 1984-85). These limits correspond to prices of 0.72 and 0.91 relative to the 1984 blended price of high quality wheats ($l86/tonne). 137 The adoption of medium quality wheat i s also constrained by agronomic considerations which are more d i f f i c u l t to detail in a model. I n i t i a l l y , the assumption was made that medium quality wheat could only compete directly with hard wheat for land allocated to wheat production. The indirect substitution of HT320 for other crops was made possible through the optimal choice of cropping ratios. The set of cropping ratios i n the model represent rotational constraints, and are based on historical observations. The model was f i r s t solved excluding medium quality wheat i n order to obtain a base case. Secondly, the new variety was introduced and the model re-solved for the eight prices of HT320 considered. The production responses for these price scenarios were compared with that of the base case to evaluate the impact of introducing medium quality wheat on production, exports, and income. Results show that although total wheat acreages increase marginally over the price range, class composition changes dramatically. The percentage of medium quality wheat increases from 10$ at $130/tonne to 94$ at $170/tonne. The adoption of the new wheat has a negative impact on the production of a l l other grains. The constraint on the amount of land available to HY320 was relaxed by allowing the new variety to compete directly with hard wheat and barley for land jointly allocated to these two crops. Once a new base case solution was obtained and the price scenarios re-run, the aggregate production effects were determined. As would be expected, more medium quality wheat i s grown at each price (an average increase of 21$). Furthermore, the negative response of feed grain and oilseed production to the price of the new crop i s greater. For the case where the medium quality wheat can only compete directly with hard wheat, an acreage response function i s constructed for each of the 138 western provinces, as well as for western Canada (Figure 4.1). The production levels of HY320 are then derived, and total and export supply functions calculated (Figure 4.2). Acreage response and supply are elastic throughout the price range, but especially when price i s low. Aggregate trends are also apparent i n the provincial results although they are less regular. At lower prices, almost a l l production of HY320 occurs i n Saskatchewan. As price increases, the distribution of medium quality wheat acreage by province approaches the distribution of wheat production i n the base case (63$ i n Saskatchewan, 23? i n Alberta, 13$ in Manitoba, and 0.5$ in B.C.). As the price rises, the new variety becomes profitable to farmers i n an increasing number of regions. The c r i t i c a l relative price for most regions i s between 0.78 and 0.83. The relative yield advantage i s greater i n the black s o i l zone than i n the brown zone, consequently medium quality wheat tends to be adopted f i r s t i n the black, dark brown, and gray s o i l zones, followed by areas i n the brown zone. When the price i s $l40/tonne, the new variety i s grown i n a band from east central Saskatchewan to west central Manitoba. At $l60/tonne, i t i s grown in a l l prairie regions except i n the brown s o i l zone of south-east Alberta and south-west Saskatchewan. The introduction of medium quality wheat affects both the level and class composition of wheat exports. As the price of the new wheat rises, the quantity of hard wheat exported declines, although total wheat exports increase. When the price i s $135/tonne, wheat exports, which are up by 5$, consist of 6$ medium quality wheat and 94$ hard wheat. At $170/tonne, wheat exports have risen by 37$, and 98$ of these exports are medium quality. The introduction of the new variety has a small but negative effect on the production and export levels of non-wheat crops. In aggregate, total grain 139 exports do increase significantly, but not at the same rate as wheat exports considered alone. Any estimate of income effect i s sensitive to assumptions regarding relative prices. Over the price range examined, the increase i n net farm income of prairie grain producers varies from $9 million to $715 million. The distribution of these gains i s largely determined by the adoption rate across regions. As would be expected, export earnings s t r i c t l y increase with price, and the gains range from $109 million to $805 million. The differences between net farm income benefits and gains i n export earnings (approximately $100 million) represent the additional costs of production and shipping, and thus includes the benefit to the transportation sector. In the context of the model, the "agricultural sector" refers to a l l those crop and livestock production a c t i v i t i e s included. The increase i n sector earnings varies from $100 million to $745 million over the set of prices considered. By subtracting the benefits to prairie grain producers, the net benefit accruing to the livestock sector and to crop production i n eastern Canada can be determined. These gains are substantial when the price of medium quality wheat i s low ($92 million at $135/tonne), but diminish as the relative price increases. The presentation of the results has been organized by production, export and income effects. Before interpreting these results, the following modifying assumptions of the model must be considered: prices of traditional grains are held constant throughout the study; the characterization of agronomic and biological constraints i s highly simplified; risk i s largely ignored; a l l grains are shipped to port at the current (subsidized) rate; and f i n a l l y , the analysis of income effects focuses on the benefits of medium quality wheat to prairie grain producers. 140 With these considerations i n mind, the policy implications of the results can be discussed. 5.2 Policy Implications of the Introduction of Medium Quality Wheat Certain policy implications can be drawn from the results of this study. These include wheat licensing arrangements, plans regarding grain storage and handling f a c i l i t i e s , policies with respect to transportation rate structures, government programs targetted at specific groups benefitting from medium quality wheat, and policies regarding the direction of future research. As mentioned earlier, there was i n i t i a l reluctance on the part of the Canadian Grain Commission to the licensing of medium quality wheat. However, in January 1985, the medium quality wheat category, "Prairie Spring Wheat", was created, and HY320 was licensed for production. Earlier studies estimated that this could result i n substantial benefits to prairie grain producers (Henning, Ulrich and Furtan, Loyns and Carter). The current study corroborates these findings but stresses the relationship between expected benefits and the ultimate price of medium quality wheat. The price realized w i l l depend on world demand and supply of medium quality wheat which may be influenced by the intervention policies of competing nations. This study estimated potential market share as a function of price. Even at relatively low prices, the resulting production level translates into a sizeable share (eg. 26$ market share at $l45/tonne), and consequently, the probability of a major competitor retaliating would be high (Alaouze et a l ) . The implication i s that either the ultimate price w i l l be lower than expected, or else Canada may be forced to consider a program of production quotas or inventory control to stay within an acceptable market share. Thus, the benefits of 141 medium quality wheat may be less than those predicted by earlier studies, and by the current analysis. If this i s the case the caution regarding the licensing of the new variety may have been jus t i f i e d . The export volumes predicted i n this study can be used to assess the impact of Canadian medium quality wheat exports on competitors and consequently provides a basis for evaluating the risk of retaliation. This information, i n conjunction with a knowledge of traditional trading partners, can assist i n developing a marketing strategy. For example, Canada may try to target i t s effort on a particular market area. The marketing strategy may also involve nnon-price n competition such as "exporting bagged wheat when necessary, being able to extend credit or subsidize interest rates, and blending wheats i n Canada before export" (Ulrich and Furtan, p. 24). Besides the potential cost of managing inventories, the increased volumes associated with the higher yielding wheat w i l l have implications for the grain delivery system. The study identifies the production pattern, and volumes expected at different relative price levels. This information can be used to assess the adequacy of the current system and changes to i t that may be of importance. This same information i s relevant to transportation rate policy. The current rate structure specifies that grain volumes i n excess of the 1981/82 level must be shipped at the compensatory rate. This policy was based on a threshhold which may no longer be appropriate with the introduction of the higher yielding wheat. However, i f the intention i s to move to a rate structure more i n line with the compensatory rate, then the current policy may be most appropriate. This study presents the regional distribution of benefits to grain 142 producers. In addition, i t identifies other beneficiaries including l i v e -stock producers, eastern grain farmers, and the transportation sector. This information could be useful in formulating government policies targetted at these groups. For example, with the growing concern over farm bankruptcies, the results can be used i n the determination of the level of assistance to a particular group or region. As a second example, during the "Crow debate" i t was argued that one effect of increasing western transportation rates would be to stimulate the prairie livestock sector by lowering the cost of feed grain. The current results indicate that the introduction of medium quality wheat should also benefit livestock producers by providing an alternate source of feed grain. To the extent that this reduces the costs of livestock production, the value added by that sector should increase proportionately. The results also have implications for future agricultural research. 1 The Nielson report points out that i n the past, an average internal rate of return of 30 to 40$ was realized by agricultural research designed to increase production where world supply was the limiting factor. However, future research and development efforts must be better directed to food markets which are becoming "increasingly demand driven, selective and internationally competitive." In addition, there must be a greater emphasis placed on accountability to the client community and not just to the Research Branch line management. In line with these recommendations, the program to develop new cultivars of medium quality wheat was a response to changing patterns of world wheat demand. Furthermore, the results of the current study estimate the level and distribution of the research benefits accruing to the clients, that i s , the primary producers. The section quoted i s from a Study Team Report to the Task Force on Program Review headed by the Honourable Erik Neilsen. 143 5.3 Extensions and Modifications of the Model Although the regional agricultural model developed i n this study i s applicable to a wide range of issues i n i t s current form, a number of suggestions w i l l be proposed to enhance i t s usefulness. Some of the proposals have direct relevance to the medium quality wheat issue while others are more general in nature. These proposals w i l l either broaden the scope of the model, or increase the detail included, thereby improving the model's analytical a b i l i t y . A number of the suggestions involve endogenizing variables which are currently specified exogenously. The proposed extensions w i l l deal with changes to the production a c t i v i t i e s f i r s t , followed by enhancements to the transportation and marketing blocks. F i r s t l y , production a c t i v i t i e s could be modified to incorporate risk, as risk factors are believed to be an important influence in farmer's decisions. One approach would be to revise the objective function so as to minimize the variance i n net returns for a given level of earnings (Hazell). This change would increase the capacity of the model to analyze questions regarding new technologies which may involve greater risk. This would require the solution of a quadratic programming problem rather than a simpler linear programming one. The approach used to model the crop sector could be modified i n several additional ways. Since wheat continues to be a dominant crop on the Prairies, the wheat categories could be differentiated into grades so that regional differences i n grading patterns could be captured. The importance of these differences to the regional analysis was stressed by Dlrich and Furtan. It may also be useful to include additional crops or crop varieties. By including soft wheat and Durum, the complete range of wheats traded internationally would be represented. Note, however, these two 144 categories account for less than 15$ of current Canadian production. As a third suggestion, the number of crop regions could be increased, especially in eastern Canada. Currently, most of the cropping detail concerns prairie crop production. This i s appropriate for the current study as medium quality wheat w i l l be grown mainly on the prairies. However, the analysis of another issue may involve the effect of a certain policy on other areas of the country, and therefore more detail may be required. The accurate modelling of agronomic considerations i s viewed to be extremely important. At present, agronomic constraints are modelled by cropping mixes, or ratios. The sensitivity of the model to these ratios was illustrated by the experiment i n relaxing these constraints i n Section 4.3.5. When wheat and barley were allowed to compete directly for land jointly available for wheat and barley production, wheat acreage increased by 11$ in aggregate and barley acreage decreased by 20$. A preferable method would be to allow the model to freely determine the optimal rotation without these relatively inflexible constraints. One approach would be to replace the set of cropping ratios by a set of planting a c t i v i t i e s based on preceding crops grown (El-Nazer and McCarl). The gross margin associated with each activity could be determined using regression analysis of historical cropping records. This method has the disadvantages of significantly increasing the size and the data requirements of the model. Furthermore, i t would be inappropriate for new varieties with limited past performance data. A second approach suggested by Howitt and Mean i s based on the penalty function concept. Since this method makes use of available information (i.e. the penalty costs i n the current solution), i t would avoid some of the limitations of the f i r s t approach. Other suggestions regarding agronomic constraints include allowing movement of land between categories 145 (e.g. the conversion of tame hay to cropland), and adding labour and equipment constraints thus permitting a more detailed consideration of the timing of operations. Carryover effects associated with such practices as the fallowing of land could be best handled i n a multi-year model. The livestock blocks could be enhanced by endogenizing the opening and closing animal numbers, especially those relating to the beef and pork sectors. This would improve the analysis of questions involving crop-livestock interactions. In addition, the number of livestock regions could be increased. Ideally, the boundaries of the crop and livestock regions would correspond resulting i n a smoother interface between the two sectors. The modelling of the transportation sector could also be revised to allow a more complex rate structure. As pointed out, current policy defines a two price system in which a subsidized rate i s charged for volumes up to a threshold level, after which the compensatory rate applies. Further refinements to the shipping block could include adding additional ports such as Prince Rupert, and increasing the modes of transportation (eg. truck, lake steamer). Finally, the marketing block could be enhanced by endogenizing inventory and domestic demand, as well as commodity prices (Duloy and Morton; Rae; Martin and Zwart). The current study has provided a model for the supply of wheat i n Canada, and i n this model, market prices have been set exogeneously. In a broader context, i t i s recognized that world prices are determined by both world supply and demand, and therefore, a general equilibrium model incorporating these relationships would be useful. Henning has developed an econometric model of the world wheat market i n which Canada's supply i s set exogeneously. As an i n i t i a l step, the current model could be used to 146 endogenize the Canadian component i n Henning's model. However, more work would s t i l l be required i n order to model market structure considerations including the potential for market intervention by major competing nations. Finally, the current study has demonstrated how the sectoral model presented can be used for policy analysis. More generally, a wide range of issues may be considered within the current structure or through modifications thereof. Some extensions to the model have been suggested in this concluding section. Although the model was designed to be f a i r l y flexible, the ease with which the various proposals can be implemented necessarily differs. The relative importance of each depends on the issue to be analyzed. However, they a l l contribute to a more powerful model for research and policy analysis. 147 BIBLIOGRAPHY Alaouze, C.M., A.S. Watson, and N.H. Sturgess. "Oligopoly Prioing i n the World Wheat Market." Amer. J. Agr. Econ. 60(1978):173-85. Canadian Grains Industry. St a t i s t i c a l Handbook 84. Winnipeg: Canada Grains Council, 1984. Canadian Wheat Board. Grain Matters. June, 1984. Carter, C., and A. Schmitz. "Import Tariffs and Price Formation i n the World Wheat Market." Amer. J. Agr. Econ. 61(1979):517-22. Colman, D. "A Review of the Arts of Supply Response Analysis." Rev. Mktg.  and Agr. Econ. 51(1983):201-30. Duloy, J.M. and R.E. Norton. "Prices and Income i n Linear Programming Models." Amer. J. Agr. Econ. 57(1975):591-600. "CHAC: A Programming Model of Mexican Agriculture." Multi-Level  Planning: Case Studies i n Mexico. ed. L. Goreux and A. Manne, pp. 291-337. Amsterdam: North-Holland Publishing Co., 1973. Economic Growth: Agriculture, a Study Team Report to the Task Force on  Program Review. Ottawa, Canada, 1985. El-Nazer, T. and B. McCarl. "Choice of Crop Rotation." Amer. J. Agr. Econ. 68(1986):127-36. Gilmour, B. "Medium Quality Wheat Production on the Canadian Prairies: Some Comments." Paper presented at Canadian Regional Agricultural Model Workshop, Ottawa, Apr. 1986. Grennes, T., P.R. Johnson and M. Thursby. The Economics of the World Grain  Trade. New York: Praeger. 1978. Hassan, Z.A. and H.B. Huff, eds. Agricultural Sector Models for Policy  Analysis. Ottawa: Agriculture Canada, 1985. Hazell, P. "Endogenous Input Prices i n Linear Programming Models." Amer. J. Agr. Econ. 61(1979):476-81. Heady, E.O. "Models for Agricultural Policy: the CARD Example." Euro.  Rev. Agr. Econ. 10(1983):1-14. Henning, J. Ph.D. thesis, School of Agricultural Economics and Extension Education, University of Guelph, 1985. Horner, W.H., J.A. Brown, J.C. Gilson, G. Lee, F.V. MacHardy, H.D. McRorle, and CL. Sibbald. Western Canadian Agriculture to 1990. Calgary: Canada West Foundation, 1980. 148 Howitt, R.E. and P. Mean. "A Positive Approach to Microeconomic Programming Models." Dept. Agr. Econ. Work. Pap. No. 83-6, University of California, Davis, 1983. Klein, K.K., CA. Webber and J.D. Graham. Medium Quality Wheat Production  on the Canadian Prairies: An Economic Assessment. Vancouver: Department of Agricultural Economics, University of British Columbia, 1986. Klein, K.K., CA. Webber and J.D. Graham. Regional Impacts from Selected  Policies on Freight Rates of Prairie Grains and Oilseeds. Vancouver: Department of Agricultural Economics, University of Briti s h Columbia, 1986. Loyns, R.M.A. and CA. Carter. Varietal Licensing Standards and Canadian  Wheat Exports. Mimeographed. Department of Agricultural Economics and Farm Management, University of Manitoba, Winnipeg, 1984. McCarl, B.A. "Model Validation: An Overview with Some Emphasis on Risk Models." Rev. Mktg. and Agr. Econ. 52(1984):153-69. McCarl, B.C. "Cropping Activities i n Agricultural Sector Models, a Methodological Proposal." Amer. J. Agr. Econ. 64(1982):768-72. McCarl, B., and T. Spreen. "Price Endogenous Mathematical Programming as a Tool for Sector Analysis." Amer. J. Agr. Econ. 62(1980):87-102. MacGregor, R.J. and J.D. Graham. "Regional Impacts of the 1985 U.S. Farm B i l l on Canada's Grain Sector: A Regional Programming Approach to Policy Analysis." Unpublished, 1986. Martin, L. and A.C Zwart. "A Spatial and Temporal Model of the North American Pork Sector for the Evaluation of Policy Alternatives." Amer.  J. Agr. Econ. 57(1975):55-66. Miketinac, J. "Technological and Institutional Change: The Case of HY320." M.Sc. Agr. thesis, Department of Agricultural Economics, University of Saskatchewan, Saskatoon, 1985. Nerlove, M. The Dynamics of Supply: Estimation of Farmer's Response to  Price. Baltimore: Johns Hopkins Press, 1985. Pachico, D., W. Janssen, and J. Lynam. "Ex ante Analysis of New Technology: A Comparison of Cassava for the Feed and Fresh Markets i n Colombia." Agricultural Systems. 11(1983):131-1^2. Rae, A.N. "Profit Maximization and Imperfect Competition - an Application of Quadratic Programming to Horticulture." J. Agr. Econ. 21(1970): 133-40. Shumway, CR. and A.A. Chang. "Linear Programming Versus Positively Estimated Supply Functions: An Empirical and Methodological Critique." Amer. J. Agr. Econ. 59(1977):344-57. 149 Thompson, K.J. and A.E. Buckwell. "A Microeconomic Agricultural Supply Model." J. Agr. Econ. 30(1979):1-11. Ulrich, A. and W.H. Furtan. An Economic Evaluation of Producing HY320  Wheat on the Prairies. Mimeographed. Department of Agricultural Economics, University of Saskatchewan, Saskatoon, 1984. Veeman, T. and M. Veeman. The Future of Grain - Canada's Prospects for  Grains, Oilseeds and Related Industries. Ottawa: Canadian Institute for Economic Policy, 1984. Walford, N. "The Future Size of Farms: Modelling the Effect of Change i n Labour and Machinery." J. Agr. Econ. 34(1983):407-16. Walker, N. and J. Dillon. "Development of an Aggregative Programming Model of Australian Agriculture." J. Agr. Econ. 27(1976):243-47. Webber, C.A., J.D. Graham, and K.K. Klein. The Structure of CRAM: A  Canadian Regional Agricultural Model. Department of Agricultural Economics, University of Brit i s h Columbia, 1986. Technical Appendix. Description of the Canadian Regional  Agricultural Model. Department of Agricultural Economics, University of British Columbia, 1986. Wicks, J.A. and I.R. Crellin. "Interactions of Technological Change i n Australian Agriculture: An Aggregative Programming Analysis." Rev.  Mktg. and Agr. Econ. 47(1979):51-61. Wilson, CF. Grain Marketing i n Canada. Winnipeg: Canadian International Grains Institute, 1979. Zentner, R.P., B.H. Sontang, and GJE. Lee. "Simulation Model for Dryland Crop Production i n the Canadian Prairies." Agricultural Systems. 3(1978):241-51. APPENDIX A. Costs of Production of Major Crops for Western Regions Table A.1: Costs of Production of Major Crops by Region i n B.C. ($/Hectare) Crop Grown on Cost i n Region 1 HQW Fallow 128.44 HQW Stubble 132.20 MQW Fallow 128.81 MQW . Stubble 132.52 Barley- Fallow 116.48 Barley Stubble 121.40 Flax Fallow Flax Stubble Canola Fallow 107.88 Canola Stubble 113.44 Other Fallow 217.70 Other Stubble 224.47 Fallow 24.54 \ 151 Table A.2: Costs of Production of Major Crops by Region in Alberta ($/Hectare) Cost of Production by Region Crop Grown on 1 2 3 4 5 6 7 HQW Fallow 96.79 131.68 148.43 128.81 154.81 148.23 128.17 HQW Stubble 98.05 140.60 155.11 143.96 165.38 155.38 131.95 MQW Fallow 97.06 131.93 148.83 129.23 155.30 148.78 128.81 MQW Stubble •98.35 140.85 155.50 144.38 165.83 155.87 132.52 Barley. Fallow 90.46 123.57 144.48 126.89 172.90 138.38 116.48 Barley Stubble 89.30 131.46 151.13 142.85 183.40 144.48 121.40 Flax Fallow 104.65 142.90 142.30 119.89 140.06 140.06 142.30 Flax Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Canola Fallow 93.92 144.26 159.92 140.80 161.60 137.64 107.88 Canola Stubble 0.0 0.0 0.0 161.45 172.97 0.0 113.44 Other Fallow 237.09 321.03 302.25 234.32 262.94 238.40 217-70 Other Stubble 248.36 327.23 307.39 249.40 270.06 243.71 224.47 Fallow 14.06 19.87 24.12 19.32 26.69 33.75 24.54 Table A.3: Costs of Production of Major Crops by Region i n Saskatchewan ($/Hectare) Cost of Production by Region Crop Grown on 1 2 3 4 5 6 7 8 9 HQW Fallow 102.64 106.18 103.36 98.52 131.51 107.41 116.09 123.75 117.27 HQW Stubble 103.21 105.76 103.66 98.69 135.07 111.47 127.73 13Q.34 127.03 MQW Fallow 102.89 106.43 103.66 98.79 131.83 107.74 116.38 127.36 117.55 MQW Stubble 103.44 106.03 103.93 98.96 135.39 111.76 128.05 133.51 127.31 Barley Fallow 97.46 100.00 98.07 91.03 128.07 104.38 108.45 124.98 118.46 Barley Stubble 96.15 97.68 95.41 88.31 126.84 104.33 114.23 130.27 127.80 Flax Fallow 105.14 108.33 104.65 0.0 140.06 112.53 119.89 142.30 140.06 Flax Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Canola Fallow 100.94 0.0 0.0 0.0 146.38 111.24 121.13 143.91 124.59 Canola Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 148.78 Other Fallow 212.46 185.60 192.14 178.23 245.12 224.66 221.40 232.37 183.59 Other Stubble 212.80 0.0 0.0 0.0 260.52 229.51 0.0 241.54 187.94 Fallow 14.46 13.89 15.64 14.36 21 .03 14.26 14.63 16.70 16.88 ro 153 Table A.4: Costs of Production of Major Crops by Region in Manitoba ($/Hectare) Cost of Production by Region Crop Grown on 1 2 3 4 5 6 HQW Fallow 137.14 134.25 156.79 158.91 164.64 157.53 HQW Stubble 149.67 145.32 166.69 172.48 174.13 163.33 MQW Fallow 137.49 134.59 162.62 165.26 165.09 157.97 MQW Stubble 150.01 145.69 172.03 178.23 174.58 163.78 Barley Fallow 126.39 125.43 147.07 151.10 147.64 143.42 Barley Stubble 134.10 130.12 150.06 154.19 155.62 145.00 Flax Fallow 139.34 135.95 146.85 152.02 162.74 154.09 Flax Stubble 0.0 0.0 156.41 160.14 0.0 0.0 Canola Fallow 153-42 143.84 169.11 172.18 171.73 159.77 Canola Stubble 0.0 0.0 180.36 180.58 0.0 0.0 Other Fallow 274.28 279.79 290.19 284.66 294.74 282.14 Other Stubble 278.73 287.25 297.48 291.58 300.15 287.30 Fallow 22.02 22.59 25.67 27.80 30.69 28.98 154 B. Yields of Major Crops for Western Regions Table B.1: Yields of Major Crops by Region i n B.C. (Tonnes/Hectare) Crop Grown on Yield i n Region 1 HQW Fallow 3.16 HQW Stubble 2.23 MQW Fallow 4.04 MQW Stubble 2.85 Barley Fallow 2.43 Barley Stubble 2.14 Flax Fallow Flax Stubble Canola Fallow 0.83 Canola Stubble 0.71 a Other Fallow 420.07 Other Stubble 420.07 Yields for 'other' crops measured i n '$/hectare' 155 Table B.2: Yields of Major Crops by Region i n Alberta (Tonnes/Hectare) Yield by Region Crop Grown on 1 2 3 4 5 6 7 HQW Fallow 2.04 2.29 2.11 2.90 3.61 3.05 2.77 HQW Stubble 1.37 2.16 2.00 2.04 2.55 2.15 1.96 MQW Fallow 2.40 2.80: 2.58 3.83 4.78 3.89 3.54 MQW Stubble 1.61 2.64 2.44 2.70 3-37 2.75 2.50 Barley Fallow 2.35 2.99 2.93 2.67 2.98 2.60 2.25 Barley Stubble 2.07 2.63 2.32 2.35 2.62 2.29 1.98 Flax Fallow 1.11 1.33 1.24 1.15 1.05 1.05 1.24 Flax Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Canola Fallow 1.04 1.18 1.08 1.12 1.18 0.92 0.79 Canola Stubble 0.0 0.0 0.0 0.92 0.97 0.0 0.67 a Other Fallow 392.89 471.96 499.14 427.48 484.32 405.24 420.07 Other Stubble 392.89 471.96 499.14 427.48 484.32 405.24 420.07 Yields for 'other' crops measured i n '$/hectare' T a b l e B . 3 : Y i e l d s o f M a j o r C r o p s by R e g i o n i n S a s k a t c h e w a n ( T o n n e s / H e c t a r e ) Y i e l d by R e g i o n C r o p Grown o n 1 2 3 4 5 6 7 8 9 HQW F a l l o w 1 .75 1.82 1.91 2 .01 2 . 0 4 1.96 2 . 2 8 2 . 1 7 2 . 2 2 HQW S t u b b l e 1.24 1 .28 1.28 1 .35 1 .43 1 .39 1.61 1 .53 1.57 MQW F a l l o w 2 . 2 4 2 . 2 6 2 . 4 5 2 . 3 6 2 . 8 2 2 . 8 4 2 . 7 1 2 . 8 7 2 . 7 9 MQW S t u b b l e 1 .59 1.59 1.64 1 .59 1 .98 2 . 0 0 1.91 2 . 0 2 1.97 B a r l e y F a l l o w 1.82 2 . 1 4 2 . 0 9 2 . 2 0 2 . 4 5 2 . 0 2 2 . 6 4 2 . 6 9 2 . 7 1 B a r l e y S t u b b l e 1 .60 1.70 1.84 1.94 1 .95 1 .78 2 . 0 9 2 . 1 4 2 . 1 5 F l a x F a l l o w 0 .80 1.00 1.11 0 . 0 1 .05 1 .07 1 .15 1 .24 1 .05 F l a x S t u b b l e 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 C a n o l a F a l l o w 0 . 8 3 0 . 0 0 . 0 0 . 0 1 .00 1.01 1 .19 1 .13 1 .18 C a n o l a S t u b b l e 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 8 7 O t h e r a F a l l o w 2 6 1 . 9 3 2 2 9 . 8 0 2 4 7 . 1 0 2 4 2 . 1 6 2 5 2 . 0 4 2 3 4 . 7 5 2 4 4 . 6 3 2 9 1 . 5 8 2 6 4 . 4 0 O t h e r S t u b b l e 2 6 1 . 9 3 0 . 0 0 . 0 0 . 0 2 5 2 . 0 4 2 3 4 . 7 5 0 . 0 2 9 1 . 5 8 2 6 4 . 4 0 Y i e l d s f o r ' o t h e r ' c r o p s measured i n ' ^ / h e c t a r e ' 157 Table B.4: Yields of Major Crops by Region i n Manitoba (Tonnes/Hectare) Yield by Region Crop Grown on 1 2 • ' 3 4 5 6 HQW Fallow 1.88 1.92 2.72 2.99 2.69 2.66 HQW Stubble 1.78 1.81 1.91 2.00 1.90 1.87 MQW Fallow 2.42 2.61 3.33 3.85 3.50 3.42 MQW Stubble 2.28 2.47 2.34 2.57 2.48 2.40 Barley Fallow 2.49 2.54 ;•:••; 2.99 2.92 2.61 2.71 Barley Stubble 2.20 2.02 2.38 2.57 2.31 2.15 Flax Fallow 0.87 s 0.96 1.02 1.12 1.06 0.95 Flax Stubble 0.0 0.0 0.92 1.00 0.0 0.0 Canola Fallow 0.90 0.95 1.00 1.05 1.00 0.92 Canola Stubble 0.0 0.0 0.82 0.87 0.0 0.0 a Other Fallow 484.32 489.26 506.56 511.50 516.44 489.26 Other Stubble 484.32 489.26 506.56 511.50 516.44 489.26 Yields for 'other' crops measured i n '$/hectare' 158 C. Area Seeded to Each Major Crop for Western Regions - Base Case Table C.1: Area of Major Crops by Region i n B.C. - Base Case (Thousand Hectares) Area i n Region Crop Grown on 1 HQW Fallow 61.27 HQW Stubble 12.71 MQW Fallow MQW Stubble Barley Fallow Barley Stubble 112.58 Flax Fallow Flax Stubble Canola Fallow Canola Stubble 67.55 Other Fallow Other Stubble 67.55 Fallow 61.27 Region Total 382.91 Table C.2: Area of Major Crops by Region in Alberta (Thousand Hectares) Area Seeded by Region Crop Grown on 1 2 3 4 5 6 7 Crop Total HQW Fallow 451.70 217-03 0.0 381.52 92.52 45.25 246.51 1434.53 HQW Stubble 35.83 785.33 260.67 419.45 0.0 49.56 219.39 1770.25 MQW Fallow 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 MQW Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Barley Fallow 0.0 360.22 34.44 0.0 0.0 0.0 0.0 394.66 Barley Stubble 102.64 0.0 400.02 782.34 539.71 456.31 465.90 2746.92 Flax Fallow 6.41 15.66 0.0 0.0 0.0 0.0 0.0 22.08 Flax Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Canola Fallow 6.41 78.31 63-19 0.0 0.0 35.56 0.0 183.47 Canola Stubble 0.0 0.0 0.0 242.15 131.07 0.0 349.43 722.65 Other Fallow 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Other Stubble 38.49 109.63 31.60 37.25 7.71 5.93 12.94 243.55 Fallow 464.53 671.22 97.63 381.52 92.52 80.81 246.51 2034.74 Region Total 1106.01 2237.41 887.55 2244.24 863.53 673.43 1540.68 9552.86 Table C.3: Area of Major Crops by Region i n Saskatchewan - Base Case (Thousand Hectares) Area Seeded by Region Crop Grown on 1 2 3 4 5 6 7 8 9 Crop Total HQW Fallow 418.65 573.27 1423.82 483.16 706.96 665.87 694.27 78.74 0.0 5074.74 HQW Stubble 298.23 352.52 163.79 7.72 632.09 491.56 315.99 619.22 688.06 3569.18 MQW Fallow 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 MQW Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Barley Fallow 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Barley Stubble 135.80 72.01 160.54 58.44 507.21 195.62 179.03 253-81 385.98 1948.44 Flax Fallow 19.40 10.29 0.0 0.0 40.58 48.91 12.79 25.38 16.78 174.12 Flax Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Canola Fallow 9.70 0.0 0.0 0.0 81.15 163.02 38.36 253-81 402.77 948.81 Canola Stubble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 134.26 134.26 Other Fallow 0.0 20.57 35.68 35.06 0.0 0.0 38.36 0.0 0.0 129.68 Other Stubble 58.20 0.0 0.0 0.0 60.87 65.21 0.0 38.07 50.35 272.69 Fallow 477.75 604.13 1459.49 518.23 828.69 877-79 783-79 357.93 419.55 6327-35 Region Total 1447.71 1632.79 3243-32 1102.61 2857.55 2507.98 2026.61 1626.96 2097.74 18579.26 161 Table C.4: Area of Major Crops by Region in Manitoba - Base Case (Thousand Hectares) Area Seeded by Region Crop Grown on 1 2 3 4 5 6 Crop Total HQW Fallow 0.0 0.0 21.22 25.05 0.0 1.30 47.58 HQW Stubble 597.14 315.56 232.69 282.78 189.49 143.72 0.0 1761.39 MQW Fallow 0.0 0.0 0.0 0.0 0.0 0.0 MQW Stubble — 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Barley Fallow 0.0 4.34 0.0 0.0 0.0 0.0 4.34 Barley Stubble 393-90 183.65 199.02 194.42 129.01 145.03 1245.03 Flax Fallow 91.21 20.14 0.0 0.0 20.16 21.75 153.26 Flax Stubble 0.0 0.0 61.76 81.01 0.0 0.0 142.77 Canola Fallow 103-49 94.00 0.0 0.0 16.13 21.75 235.37 Canola Stubble 0.0 0.0 48.04 48.61 0.0 0.0 96.64 Other Fallow 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Other Stubble 117.27 53.71 123.53 178.22 48.38 29.01 550.12 Fallow 194.70 118.48 21.22 25.05 36.28 44.81 440.56 Region Total 1497-71 789.89 707.49 835.15 439.45 407.38 4677-07 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0097201/manifest

Comment

Related Items